The Biology and Ecology of Carcharhiniform Sharks in the Gulf of Papua Prawn Trawl Fishery

ResearchOnline@JCU

This file is part of the following work:

Baje, Leontine Immoine (2019) The biology and ecology of carcharhiniform sharks in the Gulf of Papua prawn trawl fishery. PhD Thesis, James Cook University.

Access to this file is available from: https://doi.org/10.25903/461f%2D2453

The author has certified to JCU that they have made a reasonable effort to gain permission and acknowledge the owners of any third party copyright material included in this document. If you believe that this is not the case, please email [email protected] THE BIOLOGY AND ECOLOGY OF

CARCHARHINIFORM SHARKS IN THE GULF

OF PAPUA PRAWN TRAWL FISHERY

Thesis submitted by Leontine Immoine Baje BSc and BScH (University of Papua New Guinea)

For the degree of Doctor of Philosophy Centre for Sustainable Fisheries and Aquaculture College of Science and Engineering James Cook University Townsville, Australia November 2019

Statement of sources

Declaration

I declare that this thesis is my own work and has not been submitted in any form for another degree or diploma at any university or other institution of tertiary education. Information derived from the published or unpublished work of others has been acknowledged in the text and a list of references is given.

19th November 2019

______

Signed Date

Acknowledgements

Google Scholar uses the phrase “stand on the shoulder of giants” which is originally attributed to the philosopher Bernard of Charters. This saying for me sums up my journey through this PhD candidature. For this thesis to finally be put together it has taken the support and might of many giants in my life and I pay tribute to them here.

Firstly, my supervisory team - Colin, Andrew and Will. When I first walked into Colin’s office in 2015 he told me this could potentially become a PhD (it was a Masters study at the time), at that I squirmed and almost caught a flight back to PNG, it’s fair to say I didn’t see it coming and I was not mentally prepared for what eventually unfolded. A lot of things that I have done in my candidature have been way out of my comfort zone, but my supervisory team have been nothing short of fantastic, for also affording me their patience and understanding when I had to handle a child with medical needs and other personal matters. It’s been 6 years since I met Will White for the first time in Port Moresby with not a single clue of what would transpire henceforth from that day, what a journey! Colin, Andrew and Will, thank you for standing by me through it all and I have always found great support and positivity from our interactions. I remain grateful for your great contribution to my career and life, always.

My fellow fish and fisheries research lab mates - Samantha, Brooke, Ana, Madie, Sushmita, Shiori, Melissa, Stacy and Michael, it took me a while to get over initial barriers I had within myself, but I made a great group of friends who remain supportive to this day. Thank you for your friendship and constant encouragement. Huge thanks also to Jonathan Smart for helping a newbie to R get through something so daunting, for always having the answers when I had questions, even down to days before submission, much appreciated Jon! Also special thanks to Michelle Heupel and Karin Gerhart, the few chats we had when I caught up with either of you were always helpful, a great sense of strength and support. Thank you.

A word of sincere thanks also goes to the Australian Centre for International Agricultural research (ACIAR) and in particular Drs, Chris Barlow, Ann Fleming and Jes Sammut for

ii their support. By the same token I also acknowledge the Schlumberger Foundation Faculty for the Future Program, especially Eve Millon, Regina Hand and Elisabeth Forge. On both occasions of receiving the funding awards, I was pleasantly surprised and I remain grateful to both organisations for their contribution to making this PhD study come to life.

I have to make mention also of a great bunch of people I met in Townsville who became family away from family, they are: Wyn, Paul and Alistair McLennan, Freda Paiva and John Junior Flomo, Fiona and Cyan N’Drower, Miriam Supuma, Japhat and Liam. It was through Miriam’s insight and encouragement that I applied for and obtained further independent funding, one of the stars that aligned to make this PhD possible. Thank you Mimi! Also, sincere appreciation the PNG student body at JCU and the PNG wantoks in Townsville – thank you for the community spirit.

This PhD stretches borders, back on home soil I would like to acknowledge the contribution of the following people; Augustine Mungkaje – for helping to pave my path into a career in fisheries management and science. The late Augustine Mobiha- my first boss and mentor, a no-nonsense man with a knack for building capacity in young Papua New Guineans (there are not too many of these kind of people in PNG these days). Ludwig Kumoru – for your support and constant positivity, you thought me to look at my situation differently and not through the lens often used by others. Jacob Wani – I would not have applied for the JAF fellowship without your persistent encouragement, thank you for believing in me and seeing my potential. Brian Kumasi and Thomas Usu – I am blessed to have the likes of you two as part of my closest colleagues and friends. The late Luanah Yaman – I remain at a loss of your early departure from this life, I keep your example and words by me, thank you for all the encouragement over the years. Special thanks also to Benthly Sabub, Rachael Rabi and the staff of Fisheries Management Unit and colleagues and friends within the NFA.

Beyond fisheries circles I have been blessed to have the constant support of Professor Vojtech (Vojta) Novotny. My life took a different turn in the earliest days of my research career at New Guinea Binatang Research Centre, in 2006 when I first became a mother. However, Vojta did not let that get in the way of supporting me through to finishing my

iii honours and has since been a key referee in my research endeavours including this PhD. Thank you Vojta.

I acknowledge the support of my immediate and extended family from Walis Island and Korogu village in the East Sepik Province of PNG. My late grandfather Robert Yuwomu who saw the value of education even as an uneducated man by modern standards. He was a plantation labourer who made the decision to uproot his family from a remote village and move them across the country to Rabaul to give his children a better life that decision laid the foundations for my life and the opportunities I would have in future. Kokei – I pay tribute to you. Family is not always blood – Noel Spalding, thank you for your friendship and support in my early days in school up until the present time. In addition, to my inner circle, Gloria, Agilea, Andiraunga, Barbara, and Raka, thank you for our pact that has stood the test of time, and the countless chats on Facebook messenger when I needed to vent or just touch base. You remain my sisters always.

To my parents there are no words fitting to fully express my deep appreciation for all the love, care and support you have shown and continue to show to this child of yours who became a single parent rather quickly in life and needed your support even more. Mum - you are my backbone; I have come this far because of your great love and commitment to being a mum to me and extended to my children. I love you, thank you. To my siblings Rosemary, Adam and Gabriel, thank you for your support through thick and thin. Noel, Samuel, Faith- Ann and William - my four children, my anchor in life, who sometimes are the only reason I get up and show up each day, I love you, thank you for your love for mummy, and your sacrifice to put up with my long absences from home. Throughout my candidature I have seen one door open after another, I acknowledge the hand of God that has been upon my life, I give thanks and praise to the Almighty for the abundant blessings that has transpired into all this.

Dedication

In memory of the life, work and legacy of my mentors,

Augustine Malbojup Mobiha (1960 – 2019) and Luanah Koren Yaman (1980 – 2017)

Contribution of others to the thesis

Supervision

Professor Colin Simpfendorfer – Centre for Sustainable Fisheries and Aquaculture & College of Science and Engineering James Cook University

Dr Andrew Chin - Centre for Sustainable Fisheries and Aquaculture & College of Science and Engineering James Cook University

Dr William White – CSIRO, Australian National Fish Collection, Hobart, Australia

Financial Support

Australian Centre for International Agricultural Research (ACIAR) John Allwright Fellowship Award (2014)

Schlumberger Foundation Faculty for the Future Award (2017)

CSIRO

Statistical, analytical and editorial support

Professor Colin Simpfendorfer, Dr Andrew Chin, Dr William White, Dr Jonathan Smart, Michael Grant

Data

Funding for data collection

National Fisheries Authority (NFA) of Papua New Guinea, Australian Centre for International Agricultural Research and CSIRO Oceans & Atmosphere. (ACIAR; project FIS/2012/102)

Administrative oversight form ACIAR

Dr Chris Barlow, Dr Ann Fleming and Dr Jes Sammut

Administrative oversight from the NFA

Ludwig Kumoru, Leban Gisawa, Luanah Yaman, Brian Kumasi, Thomas Usu, Benthly Sabub, Philip Lens, Linus Yakwa, Vitolos Tomidi

NFA Fishery Observer data collection

Baera Nawia, Siwen Ohuesaho, Ronald Wala, Sarea Tova and Ian Tony.

Infrastructure support to store and process samples in Port Moresby

Dr Ralph Mana – University of Papua New Guinea

Laboratory assistance in Townsville

Kelsey Webber

JCU Administrative support

Debbie Berry, Jodie Wilson, Alex Salvador and Katherine Elliot

Academic and student wellbeing support

Liz Tynan, Kelly Johns and Keith Rowden

Ethics and approvals

All data collection conducted as part of this thesis was under the approval of the Australian Government through permit number IP 14007007. Data collection in Papua New Guinea was in accordance with the Fisheries Management Act (1998) and Gulf of Papua Prawn Trawl Fishery Management Plan (2008).

vii

Publications arising from this thesis

Baje, L., Smart, J.J., Chin, A., White, W.T. and Simpfendorfer, C.A., 2018. Age, growth and maturity of the Australian sharpnose shark Rhizoprionodon taylori from the Gulf of Papua. PloS One, 13(10), p.e0206581.

Baje, L., Smart, J.J., Grant, M.I., Chin, A., White, W.T. and Simpfendorfer, C.A., In press. Age, growth and maturity of the Australian blackspot shark (Carcharhinus coatesi) in the Gulf of Papua. Pacific Conservation Biology.

Other co-authored publications not directly related to this thesis

White, W. T., Baje, L., Simpfendorfer, C. A., Appleyard, S. A., Chin, A., Sabub, B., Rochel, E. & Naylor, G. J. P. 2019. Elasmobranch bycatch in the demersal prawn trawl fishery in the Gulf of Papua, Papua New Guinea. Scientific Reports, 9, 9254, 10.1038/s41598-019-45715- w.

Grant, M. I., Smart, J. J., White, W. T., Chin, A., Baje, L. & Simpfendorfer, C. A. 2018. Life history characteristics of the silky shark Carcharhinus falciformis from the central west Pacific. Marine and Freshwater Research, 69, 562-573, 10.1071/MF17163.

Smart, J. J., Chin, A., Baje, L., Tobin, A. J., Simpfendorfer, C. A. & White, W. T. 2017. Life history of the silvertip shark Carcharhinus albimarginatus from Papua New Guinea. Coral Reefs, 36, 577-588, 10.1007/s0033.

White, W. T., Baje, L., Sabub, B., Appleyard, S. A., Pogonoski, J. J. & Mana, R. R. 2017. Sharks and rays of Papua New Guinea, Canberra Australian Centre for International Agricultural Research

White, W.T., Last, P.R. and Baje, L., 2016. Aetomylaeus caeruleofasciatus, a new species of eagle ray (Myliobatiformes: Myliobatidae) from northern Australia and New Guinea. Ichthyological research, 63(1), pp.94-109, doi.org/10.1007/s1022.

viii

D'Alberto, B. M., Chin, A., Smart, J. J., Baje, L., White, W. T. & Simpfendorfer, C. A. 2016. Age, growth and maturity of oceanic whitetip shark Carcharhinus longimanus from Papua New Guinea. Marine and Freshwater Research, 68, 1118-1129, 10.1071/MF16165.

Smart, J. J., Chin, A., Baje, L., Green, M. E., Appleyard, S. A., Tobin, A. J., Simpfendorfer, C. A. & White, W. T. 2016. Effects of including misidentified sharks in life history analyses: A case study on the grey reef shark Carcharhinus amblyrhynchos from Papua New Guinea. PloS one, 11, e0153116.

White, W. T., Appleyard, S. A., Sabub, B., Kyne, P. M., Harris, M., Lis, R., Baje, L., Usu, T., Smart, J. J. & Corrigan, S. 2015. Rediscovery of the threatened river sharks, Glyphis garricki and G. glyphis, in Papua New Guinea. PloS one, 10, e0140075.

Abstract

The elasmobranch fauna of Papua New Guinea (PNG) and its interaction with fisheries has been poorly studied in the past. Fisheries generally adversely impact elasmobranchs due to their low productivity life histories. Without fishery and region specific data on elasmobranchs the impact on their populations cannot be fully understood and subsequent development of appropriate fisheries management and conservation measures cannot be achieved. The objectives of this thesis were to address some of these data gaps for the Gulf of Papua Prawn Fishery (GoPPF) in PNG through the assessment of biological and ecological parameters of species caught as bycatch and the development of an ecological risk assessment for all elasmobranch species caught in the fishery.

The ecological component of this work focused on the feeding relationships among the Australian blackspot shark Carcharhinus coatesi, the milk shark Rhizoprionodon acutus and the Australian sharpnose shark Rhizoprionodon taylori. Rhizoprionodon acutus had a more specific diet compared to the other species, feeding almost exclusively on teleosts while C. coatesi and R. taylori had more diverse diets that had greater overlap. The limited sampling in this study did not fully characterise the diets of the three species, however, it does provide the first empirical evidence of trophic relationships between these sympatric sharks and their prey for the Gulf of Papua.

The biology of R. taylori and C. coatesi was investigated through determination of their age, growth and maturity. Ages were determined from vertebrae samples. Length at age data were fitted to several models in a multi-model information theoretic approach to determine which model provided the best fit. Maturity was analysed using logistic regression of maturity categories recorded from samples combined with size and age data. These studies provide an understanding of the growth rate and pattern of each species and the length and age which males and females of each species reach reproductive maturity.

To assess the biology of R. taylori, 186 samples were collected comprising 131 females (31- 66 cm TL) and 55 males (31-53 cm TL). The lack of small individuals close to the size at

x birth made fitting of growth curves more difficult, two methods (fixed length at birth and additional zero aged individuals) accounting for this were trialled. The von Bertalanffy growth model provided the best fit to the data when used with a fixed length-at-birth (L0 = 26 -1 cm TL). Males (퐿∞= 46 cm TL, k = 3.69 yr , L50 = 41.7 cm TL and A50 = 0.5 years) grew at a faster rate and matured at smaller sizes and younger ages than females (퐿∞ = 58 cm TL, k = -1 1.98 yr , L5o = 47.0 cm TL and A50 = 0.93 years). However, none of the methods to account for the lack of small individuals fully accounted for this phenomenon, and hence the results remain uncertain. Despite this, the results reaffirm the rapid growth of this species and suggest that the Gulf of Papua population may grow at a faster rate than Australian populations. Rhizoprionodon taylori is possibly well placed to withstand current fishing pressure despite being a common bycatch species in the GoPPF. However, further research needs to be undertaken to estimate other key life history parameters to fully assess the population status of this exploited shark species.

Carcharhinus coatesi is a similar small bodied coastal shark to R. taylori but some differences were observed in its growth and maturity parameters. The von Bertalanffy growth model also fit the data best for C. coatesi; parameters were L0 = 40.6 cm ± 0.8, L∞ = 74.8 cm ± 2.1, k = 0.33 year1 ± 0.06. Length-at-maturity analysis indicated that males reach maturity at L50 = 66.3 cm (CI: 63.8, 71.4) and L95 = 71.6 (C1: 64.6, 74.2) cm while females matured at

L50 = 71.4 cm (CI: 61.5, 72.01) and L95 = 72.5 cm (CI: 62.7, 74.0). Age-at-maturity estimates showed that both males (A50 = 5.1 years (CI: 4.6, 7.1), A95 = 6.4 years (CI: 5.1, 7.2) and females (A50 = 5.3 years (CI: 3.5, 8.7) and A95 = 7.4 (CI: 3.6, 8.8) years) reach maturity at about the same age, but in comparison to other small bodied carcharhinids, C coatesi has slower growth in early life stages and reaches maturity at a later age. This biological trait along with a small litter size indicates that the population of C. coatesi in the Gulf of Papua may be more susceptible to decline as a result of fishing.

An Ecological Risk Assessment (ERA) was conducted to estimate the susceptibility of species caught in the fishery and the potential for a species to recover from population declines due to fishing if they occur. Of the 39 elasmobranch species encountered as bycatch in the fishery 10 were classified as being at low risk, 26 subjected to medium risk and 3 at high risk. The species at high risk were the Australian blackspot shark C. coatesi, the

xi eyebrow wedgefish Rhynchobatus palpebratus and the blackspotted whipray Maculabatis astra. This is the first ERA conducted for this fishery. The findings provide fishery managers with information to implement an ecosystem-based approach to managing the fishery to reduce bycatch and improve the sustainability of the GoPPF.

This thesis has provided new information on the diet, age, growth, maturity and the potential risk of species suffering population declines from being caught in the GoPPF. These outcomes have implications for fisheries management and conservation of species in PNG and the surrounding regions. The areas of study begin to address current data gaps for this fishery and also set the foundation for future work to improve fisheries management and protect the survival of species through conservation measures in PNG.

xii

Table of contents

Statement of sources ...... i

Acknowledgements ...... ii

Dedication ...... v

Contribution of others to the thesis ...... vi

Publications arising from this thesis ...... viii

Other co-authored publications not directly related to this thesis ...... viii

Abstract ...... x

Chapter 1 General Introduction ...... 1

Chapter 2 Resource partitioning in elasmobranchs: a review ...... 5

2.1 Introduction ...... 5 2.2 Article selection and outcome ...... 7 2.3 Development of methodologies to study resource partitioning ...... 21 2.4 Dietary partitioning ...... 24 2.5 Spatial and temporal partitioning ...... 26 2.6 Interrelationship between resource use axes ...... 27 2.7 Future work ...... 28 Chapter 3 Dietary overlap of carcharhinid sharks in the Gulf of Papua ...... 30

3.1 Introduction ...... 30 3.2 Methods ...... 32 3.2.1 Study site ...... 32 3.2.2 Sampling and sample preservation ...... 33 3.2.3 Dietary indices...... 34 3.2.4 Dietary overlap ...... 34 3.2.5 Multivariate analysis ...... 35 3.3 Results ...... 36 3.3.1 Size ranges and sample size ...... 36 3.3.2 Main prey types and proportions in diet...... 38 3.3.3 Diet Overlap ...... 41 3.3.4 Multivariate analyses...... 42 3.3.4.1 Intraspecific dietary comparison ...... 42

3.4 Discussion ...... 48 Chapter 4 Age, growth and maturity of the Australian sharpnose shark Rhizoprionodon taylori from the Gulf of Papua ...... 51

4.1 Introduction ...... 51 4.2 Materials and methods ...... 53 4.2.1 Vertebrae preparation ...... 57 4.2.2 Age determination ...... 57 4.2.3 Partial ages ...... 58 4.2.4 Growth model fitting ...... 59 4.2.5 Maturity ...... 63 4.3 Results ...... 63 4.3.1 Age determination ...... 63 4.3.2 Growth model fitting ...... 66 4.3.3 Maturity ...... 73 4.4 Discussion ...... 74 Chapter 5 Age, growth and maturity of the Australian blackspot shark Carcharhinus coatesi in the Gulf of Papua...... 81

5.1 Introduction ...... 81 5.2 Materials and methods ...... 83 5.2.1 Vertebrae preparation ...... 85 5.2.2 Age determination ...... 85 5.2.3 Precision and bias ...... 87 5.2.4 Growth model fitting ...... 87 5.2.5 Maturity ...... 89 5.3 Results ...... 90 5.3.1 Age determination ...... 90 5.3.2 Growth model fitting ...... 91 5.3.3 Maturity ...... 94 5.4 Discussion ...... 98 Chapter 6 Ecological Risk Assessment of elasmobranchs caught in the Gulf of Papua prawn fishery ...... 103

6.1 Introduction ...... 103 6.2 Methods ...... 105 6.2.1 Sample collection ...... 105 6.2.2 Ecological Risk Assessment...... 106

6.2.2.1 Susceptibility of being caught by the fishery ...... 106 6.2.2.2 Recovery potential...... 107 6.2.3 Analysis of criteria ...... 109 6.3 Results ...... 112 6.3.1 Elasmobranch species encountered in the fishery ...... 112 6.3.2 Assessment of risk ...... 120 6.4 Discussion ...... 123 Chapter 7 General Discussion ...... 127

7.1 Future research ...... 134 References ...... 135

List of Tables

Table 2-1: List of studies on elasmobranch resource partitioning ...... 8

Table 3-1: Percent frequency of occurrence (% FO) and percent by number (% N) of prey categories found in the stomachs of Rhizoprionodon taylori, Rhizoprionodon acutus and Carcharhinus coatesi in the Gulf of Papua...... 40

Table 4-1: The maturity of male and female samples were determined by the state of the uteri and ovaries in females, and claspers in males. Maturity stages were assigned a binary category for statistical analysis...... 56

Table 4-2: Equations of the three growth functions used in the multi model approach ...... 60

Table 4-3: Summary of results from the multi model approach incorporating Akaike’s information Criterion (AIC) using three-parameter versions of models...... 68

Table 4-4: Summary of results from the multi model approach incorporating Akaike’s information Criterion (AIC) using three-parameter versions of models with four hypothetical aged zero individuals...... 69

Table 4-5: Summary of results from the multi model approach incorporating Akaike’s information Criterion (AIC) using two parameter versions of growth models with a fixed length-at-birth for Rhizoprionodon taylori from the Gulf of Papua...... 71

Table 5-1: The maturity of male and female samples were determined by the state of the uteri and ovaries in females, and claspers in males. Maturity stages were assigned a binary category for statistical analysis...... 84

Table 5-2: Equations of the three growth functions used in the multi model approach ...... 89

Table 5-3: Summary of results from the multi model inference framework (MMI) incorporating Akaike’s Information Criterion (AIC). n is the sample size, AICC is the small- sample bias adjusted Akaike’s Information Criteria, ∆ is the difference in AICC values between models, w (%) are the AICC weights, L0 and L∞ are the length-at-birth and asymptotic length in cm respectively, k is the growth completion rate in (year-1) for the

VBGM, g(log) and g(gom) are the growth parameters for Logistic and Gompertz functions respectively, SE is the standard error of each growth parameter and RSE is the residual standard error for the model...... 94

Table 6-1: The criteria used to assess and assign ranks for each criteria considered under susceptibility and the potential to recover from the effects of fishing...... 110

Table 6-2: Range of Euclidean distances and corresponding level of risk...... 111

Table 6-3: Elasmobranch species taken as bycatch in the Gulf of Papua prawn trawl fishery...... 113

Table 6-4: The abundance, percentage of total elasmobranch catch, size range and maximum known size of species caught in the GoPPF from all sampling trips...... 115

Table 6-5: Partial correlation of criteria used in the study...... 119

Table 6-6: The Euclidean distance from the point of lowest risk to each species, the associated level of risk and the IUCN status of elasmobranch bycatch caught in the GoPPF...... 121

List of Figures

Figure 3-1: The Gulf of Papua situated in the south of Papua New Guinea...... 33

Figure 3-2: Length frequency of Carcharhinus coatesi, Rhizopriondon acutus and Rhizoprionodon taylori caught in the Gulf of Papua prawn trawl fishery and used for stomach content analysis ...... 37

Figure 3-3: Cumulative prey curves for Carcharhinus coatesi (blue) Rhizoprionodon acutus (green) and Rhizoprionodon taylori (red) from the Gulf of Papua...... 38

Figure 3-4: non-Metric multidimensional scaling ordination of dietary composition by number (%N) of Rhizoprionodon taylori, Rhizoprionodon acutus and Carcharhinus coatesi in the Gulf of Papua...... 43

Figure 3-5: non-Metric multidimensional scaling ordination of dietary composition by number (%N) according to North-west monsoon and South-east monsoon periods that occur in the Gulf of Papua...... 44

Figure 3-6: Composition of the diets of Carcharhinus coatesi, Rhizoprionodon acutus and Rhizoprionodon taylori according to different size classes...... 46

Figure 3-7: non-Metric multidimensional scaling ordination of dietary composition by number (%N) of size classes of Rhizoprionodon taylori (1), Rhizoprionodon acutus (2) and Carcharhinus coatesi (3) in the Gulf of Papua...... 47

Figure 4-1: The Gulf of Papua is situated along the southern coast of Papua New Guinea. The insert shows the distribution of Rhizoprionodon taylori in Australia...... 55

Figure 4-2: Frequency histogram of samples for each age class...... 65

Figure 4-3: Age bias plot showing agreement between two independent readers. The PA ± 1 year was 62.4%, APE was 29.1 and Chang’s coefficient of variation (CV) was 41.1%...... 66

Figure 4-4: Two (VB2) and three parameter (VB3) length-at-age curves for female and male Rhizoprionodon taylori from the Gulf of Papua fitted with 95% bootstrapped confidence intervals...... 73

Figure 4-5: Age and length-at-maturity ogives for female and male Rhizoprionodon taylori from the Gulf of Papua. The large points on the curve represent the length and age at which

50% of population reaches maturity. 95% bootstrapped confidence intervals are indicated with shaded areas...... 74

Figure 5-1: Cross-section of a Carcharhinus coatesi vertebral centrum viewed under a microscope. Birthmark and annual band pairs indicate, 8 years of age...... 86

Figure 5-2: Age bias plot showing agreement between two independent readers. The percentage agreement_1 year was 58.7%, Average Percentage Error was 9.93% and Chang’s coefficient of variation was 14.05%...... 91

Figure 5-3: Length-at-age curve for Carcharhinus coatesi from the Gulf of Papua with both sexes combined fitted with a three-parameter von Bertalanffy growth model (solid line) and 95% bootstrapped confidence intervals (dotted lines)...... 93

Figure 5-4: Length-at-maturity ogives for (a) male and (b) female Carcharhinus coatesi from the Gulf of Papua. The shaded points represent the length at which 50% of the population reaches maturity. The 95% confidence intervals are indicated with dashed lines...... 95

Figure 5-5: Age-at-maturity ogives for (a) male and (b) female Carcharhinus coatesi from the Gulf of Papua. The shaded points represent the ages at which 50% of the population reach maturity. The 95% confidence intervals are indicated with dashed lines...... 96

Figure 5-6: Age frequency of individual Carcharhinus coatesi sampled. The dotted lines indicated age-at-maturity for (a) males and (b) females...... 96

Figure 5-7: von Bertalanffy growth curves of small-bodied carcharhinids. Data sources: Rhizoprionodon terraenovae (Loefer and Sedberry 2003), Rhizoprionodon acutus (Harry et al. 2010), Carcharhinus coatesi (Aus) (Smart et al. 2013), Loxodon macrorhinus (Gutteridge et al. 2013), Rhizoprionodon taylori (Simpfendorfer 1993), Scoliodon laticaudus (Nair 1976)...... 97

Figure 6-1: Ecological Risk Assessment of elasmobranch species caught in the GoPPF based on the susceptibility of each species to the fishery and their estimated potential to recover from the effects of fishing...... 120

Chapter 1 General Introduction

Elasmobranchs (sharks and rays) are a subclass of Chondrichthyes (cartilaginous fishes) that have lived for more than 420 million years (Simpfendorfer and Dulvy, 2017). Elasmobranchs inhabit a variety of habitats but the majority of species are marine dwelling (Musick et al.,

2004). As predators, sharks and rays are functional members of food chains that impede the population growth of their prey through direct feeding and by inducing anti-predator behaviour (Heithaus et al., 2008). The effect on prey communities is dependent on size, where large sharks in particular with few natural predators function as apex feeders with a strong top down influence on prey populations (Stevens et al., 2000). Smaller sized sharks and rays are regarded as meso-predators that provide an intermediate link between top and lower trophic levels (Kinney et al., 2011, Heupel et al., 2014).

In recent times, many elasmobranch populations have faced drastic population declines due to anthropogenic factors, in particular increasing global fisheries (Dulvy et al., 2014). The vulnerability of these fishes stems from the inability of their populations to regenerate within short time frames to compensate for the consistent loss of individuals harvested through fishing (Herndon et al., 2010). This is attributed to biological traits such as slow growth, late maturity and small litter sizes that are common among elasmobranch species (Stevens et al.,

2000, Smith et al., 1998). It is now estimated that up to a quarter of all known elasmobranch species face some level of threat to their survival (Dulvy et al., 2014). Therefore there is growing international concern focused on the protection of these species (Camhi et al., 2009)

1 and the possible ecological ramifications of the absence of strong predatory roles that elasmobranchs exert in aquatic food webs (Grubbs et al., 2016).

Improving the sustainability of fisheries is an important step in addressing the steady decline in elasmobranch populations. This requires fishery specific data on catches of sharks and rays and information about their taxonomy, life history and ecology. In particular, life history information is crucial to fisheries management because it provides fundamental biological characteristics such as the growth rate, size and age at first maturity, periodicity of reproduction, the number of young produced, and maximum age (Gallucci et al., 2006). Such information then forms the basis for wider demographic analyses which can estimate the risks of extinction (Pardo et al., 2016b) and be used in ecological risk assessments where full scale stock assessments are lacking (Braccini et al., 2006). Noting that life history traits can differ among populations of the same species in different localities (Lombardi-Carlson et al., 2003,

White, 2007a) it is imperative to determine population and region specific parameters to provide more accurate management advice. An understanding of the ecological traits of species on the other hand further enhances knowledge of ecosystem processes (Barría et al.,

2015) and is also a precursor for assessing the level of risk a fishery poses to vulnerable species. Ecological species information is vital for managing fisheries from an ecosystem perspective which is a management approach in line with modern practices of responsible fisheries (Astles et al., 2006).

The acquisition of fishery-specific data varies worldwide but generally many fisheries remain data poor despite many years of operation especially pertaining to bycatch (Barker and

Schluessel, 2005). Although there are fisheries based on commercially valuable sharks and

2 rays, a large portion of elasmobranch mortality is often caught incidentally and regarded as bycatch (Dulvy et al., 2008, James et al., 2016) . These are often ignored in standard data collection practices (Clarke et al., 2013), which means that in many instances fisheries managers do not know what species are caught and the rates at which they are exploited.

Given the vulnerability of elasmobranch life histories it is possible that in regions of high fishing pressure many elasmobranch populations may have been overfished without detection. The lack of quantitative information on which to base fisheries management decisions is more severe in developing countries (Fuentes and de Leon Corral, 1997) and data collection in these regions will have potentially significant benefit for developing future management actions.

Papua New Guinea (PNG) is a developing nation located in a region of high marine biodiversity (Allen, 2008), including elasmobranchs (White et al., 2017b). In PNG, elasmobranchs have cultural, food security and socio-economic value through fisheries.

Sharks and rays support livelihoods through subsistence and semi-economic fishing in coastal communities across 14 maritime provinces. Coastal fishers actively target elasmobranchs for their fins to be sold to domestic buyers and eventually reach international markets (Sabetian and Foale, 2006) while the meat from some more palatable species is sold locally. Sharks were also targeted in a dedicated shark longline fishery that ceased operation in mid-2014

(Usu et al., 2015, Smart et al., 2016a). In addition, a large proportion of elasmobranchs are harvested as bycatch in the industrial tuna longline, purse seine and trawl fisheries. Despite the varied uses of elasmobranch resources in PNG, there remains a paucity of research

(White et al., 2015) to underpin improving and strengthening fishery management practices for the stocks that are impacted. In an effort to address this severe lack of information, recent research has been conducted on elasmobranch life history (Smart et al., 2017a, D'Alberto et

3 al., 2016), demography (Smart et al., 2017b) and population connectivity (Green et al., 2019) for key species encountered mainly in the tuna and shark longline fisheries. However, there is a need for similar work to be done with respect to the coastal and trawl fisheries.

The Gulf of Papua Prawn Fishery (GoPPF) is currently the only trawl fishery operating in

PNG, supplying domestic and international markets with wild caught prawns. Trawling operations are concentrated in the inshore areas of the Gulf of Papua (GoP) and capture a variety of bycatch in the trawl nets, including elasmobranchs. Despite a 40 year history of the fishery, the species composition of elasmobranch bycatch has only recently been established

(White et al., 2019). Trawl fisheries, particularly in regions of high bio-diversity such as

PNG, impact a wide range of species due to the gear type and the method of fishing used

(Oliver et al., 2015). This scenario, alongside poor data collection on bycatch, leaves scarce information to assess the sustainability of impacted populations. Therefore, this research was aimed at further advancing investigations into the biology and ecology of common bycatch species, and the sustainability of elasmobranch bycatch in the GoPPF. In order to achieve these aims, this thesis comprises a literature review and four main data chapters: Chapters 2 and 3 focus on the ecological component of this thesis. Chapter 2 is a literature review on research that has contributed to the understanding of resource partitioning in elasmobranchs while Chapter 3 explores the dietary overlap of three sympatric shark species. Chapters 4 and

5 examine life history of two of the three key shark species that are most commonly captured in the GoPPF and were featured in Chapter 3; and to investigate the effect of the fishery on the entire range of elasmobranch bycatch in the fishery, Chapter 6 is an ecological risk assessment to identify species that are most at risk from impact of commercial prawn trawling in the Gulf of Papua.

Chapter 2 Resource partitioning in elasmobranchs: a review

2.1 Introduction

The ecological niche of one organism overlaps with another when they use the same resources that are available to them in the environment. The interaction becomes competitive when shared resources are limited (Krebs, 2009) which can lead to the decline in population size and eventual extinction of a species (Schoener, 1974). A theory originally postulated by

Gregory Gause implies that species cannot co-exist for long periods if they use very similar resources (Schoener, 1974). However, in nature great numbers of species are able to co-exist and seemingly share resources. Resource partitioning is recognised as a mechanism used to circumvent the harmful consequences of competition and enable co-existence, maintaining biodiversity in terrestrial and aquatic systems (Chesson, 2000). In general, the dimensions across which the use of resources can be divided are food, habitat, and time. An early review of resource partitioning studies on fish revealed that partitioning is most likely to occur by trophic differences, followed by space use and finally temporal separation (Ross, 1986).

Earlier studies of resource partitioning conducted on fish species were based on teleost species and largely ignored elasmobranchs (sharks and rays) (Ross, 1986) though they play a key role in shaping the structure of marine ecosystems (Heithaus, 2004). Their ecological roles as top-level and meso-predators is understood on a general level but are still being defined in variable contexts (Heupel et al., 2014). Furthermore, the direct feeding interactions of elasmobranchs and the indirect consequences of their presence such as inducing behavioural changes in prey species is yet to be fully investigated (Heithaus et al., 2008). The decline of elasmobranch populations due to various anthropogenic effects can lead to

5 ecosystem shifts in the diversity and abundance of marine species which results in further changes in the food web. Over the past three decades there has been a gradual increase in the number of studies focused on the ecology of elasmobranchs which has also integrated assessments of resource partitioning. Ultimately these studies provide an understanding of the role of elasmobranchs in ecosystems (Simpfendorfer et al., 2011) and enable changes in communities as a result of declining elasmobranch populations to be predicted (Shipley et al.,

2018).

Knowledge of elasmobranch resource use patterns aids in conservation planning and the management of fisheries from an ecosystem perspective (Munroe et al., 2014, Bethea et al.,

2004). Studies have focused on dietary, spatial and temporal partitioning using a range of methods to focus on a single species, sympatric elasmobranchs, or groups of co-occurring elasmobranch and non-elasmobranch species. The extent of resource use overlap in food consumption or space use indicates the level of competition for resources. While this body of work has contributed to a more in-depth understanding of elasmobranch interactions across the various resource use planes individually, it has also revealed that the ecology in this group of fishes is dynamic and resource partitioning across all scales are interrelated to maintain co- existence of species. This chapter reviews the existing literature on resource partitioning in elasmobranchs with the following objectives: (1) describe resource partitioning identified among elasmobranchs (2) outline the methods used in studies (3) identify gaps and avenues for future research.

2.2 Article selection and outcome

Key words were used to identify studies that directly examined the occurrence of resource partitioning or inferred resource partitioning from characterisation of diets or spatial use patterns. The key search terms included elasmobranch, shark, rays and partitioning. Criteria for including a paper in this review included work that examined the diets, habitat use and seasonal patterns of habitat use of elasmobranchs as well as studies that discussed the main methodology used in fish resource partitioning studies. A total of 56 studies were identified

(Table 2-1). The greatest proportion (38%) of studies were focused on sharks. Studies on rays made up only 24% of the included papers, and the remainder of studies focused on co- occurring sharks and rays (23%) and elasmobranchs and other sympatric species (16%). The majority of studies focused on characterising trophic traits of species and dietary partitioning.

Habitat information was dominated by the determination of nursery areas and space use within these sites. The occurrence of partitioning or potential for it was recorded in the majority of studies with a minimal number of investigations not finding any evidence.

Table 2-1: List of studies on elasmobranch resource partitioning

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Heupel et al. North East Coast Examine space use, movement and Acoustic Tropical 6 Spatial (2019) Australia habitat use of sharks in a coastal telemetry nearshore partitioning bay to determine niche partitioning. embayment

Shiffman et al. Florida Hypothesise that similar prey Stable isotope Bay and reef 11 Dietary (2019) availability and overlap of study analysis partitioning species ranges the relative isotopic between two niche area and core isotopic niche species at one overlap and other trophic habitat but not interactions would remain constant observed at a across different habitats. neighbouring different habitat where both species were also present.

Curnick et al. British Indian Examine resource partitioning and Stable isotope Reef 2 Temporal (2019) Ocean Territory seasonal variation in resource use. analysis dietary partitioning.

Shipley et al. Bahamas Hypothesise that resource Stable isotope Benthic 1 2 Possible (2018) partitioning of food enables co- analysis dietary or food occurrence of species in relatively partitioning high abundance which will be among rays.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited shown in a diverse range of trophic resource pools.

Valls et al. (2017) Mediterranean Investigate the trophic relationship Stomach content Deep sea 5 18 cephalopod Food of elasmobranch and cephalopods and stable partitioning

to determine feeding strategies and isotope analysis. resource partitioning.

Gallagher et al. Southeastern Investigated inferred trophic Stable isotope Bay 3 No resource (2017) USA position, isotopic niche overlap and analysis, partitioning in patterns of resource use and estimates of prey rich area

compared this to abundance abundance and but possible information from the study area. occurrence from selective empirical shark feeding surveys overtime.

Bangley and North Carolina Investigate habitat use Multi-gear Estuarine 9 4 Temporal Rulifson (2017) USA incorporating set time in a fishery - fishery partitioning independent sampling to also focus independent among three on transition from diurnal to surveys. species and nocturnal periods to account for spatial potential temporal effects on habitat partitioning preferences and resource among two partitioning among elasmobranchs. species.

Gracan et al. continental shelf To determine the diet, feeding Fishery Continental 2 Seasonal (2017) of the Adriatic ecology and trophic position. dependent shelf dietary Sea sampling, sharks partitioning collected from

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited commercial trawlers.

Matich et al. Florida Investigate if abundant large bodied Stable isotope Estuary 1 alligators, 2 Potential (2017a) Everglades species (aquatic reptiles, analysis, teleosts and dietary and elasmobranchs, marine mammals acoustic bottlenose seasonal and teleost fishes) partition telemetry, visual dolphins habitat use. resources and habitat in low surveys, productivity environment. published diet and life history demographic information.

Matich et al. French Test the hypothesis that niche Stable isotope Coral reef 2 Dietary (2017b) Polynesia segregation occurs in response to analysis. lagoon partitioning potential completion for food only when both resources between co-occurring species are shark species within their nursery present one habitats. area.

Amariles et al. Colombian Examine feeding habits and trophic Stomach content Coastal 2 Dietary (2017) Pacific relationships between species and analysis region partitioning evaluate trophic overlap or partitioning.

Estupinan-Montano Malpelo Island, Describe trophic ecology. Stable isotope Coastal 2 Dietary and et al. (2017) Colombia analysis habitat partitioning.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Shaw et al. (2016) South Carolina, Examine the trophic ecology of Stable isotope Bay or 6 1 3 teleost Dietary USA estuarine dependent predatory fish, analysis estuary area partitioning analyse dietary niche overlap and infer potential prey.

Pardo et al. (2015) Moreton Bay, Investigate dietary partitioning Stomach content Intertidal 3 Dietary Australia between species. analysis flat partitioning

Raoult et al. (2015) Tasmania, Test the hypothesis that species will Stable isotope Coastal 2 Dietary Australia feed on different prey at different analysis partitioning - trophic levels to reduce overlap and species feed at competition despite similar different morphology. trophic levels.

Varghese et al. India, Arabian To understand prey species Stomach content Pelagic 2 1 9 other Temporal (2014) Sea composition, trophic level, diet analysis pelagic partitioning overlap and trophic organisation. species though diurnal and nocturnal feeding to avoid competition

Szczepanski and Delaware Bay, Examine diet and analyse how Stomach content 1 Ontogenetic Bengtson (2014) USA feeding habits may change analysis and, dietary temporally and ontogenetically. fisheries partitioning independent trawl surveys.

Bornatowski et al. Brazil Analyse and compare the diet of co- Stomach content Coastal bay 4 Dietary (2014b) occurring species. analysis from partitioning fishery landings. through differing diets.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Kiszka et al. (2014) Southwestern Assess the trophic relationships, Stable isotope Coastal 3 1 Possibly diet Madagascar isotopic niche breath and overlap as analysis and habitat well as ontogenetic variation in partitioning trophic interactions. also by size and sex.

Tillett et al. (2014) Northern Determine degree of dietary Stomach content Gulf area 3 Possible spatial Australia overlap, differences in type and analysis and temporal proportion of prey among species, partitioning determine if prey type increases among with maturity and correlates with juveniles, and adult feeding patterns and analyse if possible dietary dietary partitioning occurs by sex. differences.

O'Shea et al. (2013) Western Feeding biology and dietary Stomach content Coral reef 5 Diet Australia preferences were examined in order and sediment lagoon partitioning for to determine if sympatric species analysis. one species, the partition diets according to the other four degree of overlap in occupied species may habitats, so that species with partition differences in spatial and temporal resource use by scales will have the lowest diet time and space overlap. (habitat).

Taylor and Bennett Moreton Bay, Identify species composition and Commercial Bay 13 Temporal (2013) Australia sex ratio to determine size at birth gillnet catches partitioning and timing of parturition. To and fishery identify size related patterns of independent occurrence and determine extent to surveys. which shark assemblage varies among months and seasons.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Tilley et al. (2013) Belize Determine the ecological niche of a Stable isotope Reef atoll 2 1 Dietary species and compare it to other analysis partitioning sympatric elasmobranchs. among two shark species.

Heithaus et al. Shark Bay, Investigate trophic positions and Stable isotope Inshore bay 11 1 1 (dolphin) Trophic (2013) Western isotopic niches, overlap of isotopic analysis resource Australia niches among species, relationship partitioning between body size and relative based on non- trophic position and possibility for overlapping individual dietary specialisation. stable isotope results.

Navarro-Gonzalez Mexico Describe and analyse feeding habits Stomach content 6 et al. (2012) and dietary similarities. analysis

Vaudo and Western Investigate residency patterns of Acoustic Shallow 4 No apparent Heithaus (2012) Australia species and diel-patterns of habitat telemetry mud flat in temporal or use to test the hypothesis that bay. spatial individuals spend more time in partitioning. near-shore microhabitats, left the sandflats during the night and spent less time on the sand flats during the cold seasons.

Barbini and Southwestern Analyse inter-specific relationship Stomach content Coastal 2 Both species Lucifora (2012) Atlantic in diets between two analysis zone exhibited morphologically similar species. different dietary traits and therefore partitioned food resources.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Rogers et al. (2012) South Australia Examine, quantify and compare Stomach content Pelagic 5 Potential diets of pelagic sharks in a analysis dietary continental shelf region and partitioning. adjacent gulf regions. Specialised diet in one species.

Jacobsen and Northeastern Describe diets and examine intra Stomach content Coastal 3 Possible Bennett (2012) Australia and inter specific differences in diet analysis temperate dietary compositions. and tropical partitioning waters. linked to differences in prey proportion among species, size and morphology.

Yick et al. (2011) Australia Describe the diet and assess the Stomach content Bay 2 Dietary level of dietary resource analysis partitioning partitioning. through specialisation on different prey taxa despite feeding on similar prey.

Dale et al. (2011) Kanehoe Bay, Quantify the foraging ecology and Stomach content Marine bay 1 1 Dietary Hawaii habitat use and evaluate the extent analysis and partitioning of ecological interactions. bulk amino acid stable isotope analysis.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Speed et al. (2011) Northwestern Hypothesise that aggregations are Acoustic Inshore bay 4 No evidence of Australia due to reproductive purposes and telemetry and partitioning. use of site as nursery area, species visual census. and size classes partition by space use and time, species use site as a refuge through diel patterns of attendance and species show long term site fidelity.

Vaudo and Shark Bay, Examine the trophic niches of a Stomach content Marine bay 2 11 No clear Heithaus (2011) Western nearshore elasmobranch and stable evidence of Australia community. isotope analyses. differences in diets but attribute co- existence and high diversity to individual specialisation.

Sommerville et al. Southwestern Examine similarities in diets of Stomach content Coastal 1 3 Diets change (2011) Australia species and how this change with analysis region seasonally and body size and season. intra-and inter- specific differences in diets facilitate co-existence through food partitioning.

Kinney et al. Cleveland Bay, Assess the extent of dietary Stable isotope Bay 7 3 teleost Dietary (2011) QLD, Australia partitioning within and established analysis species partitioning shark nursery. occurring among species

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited in the communal nursery.

Taylor et al. (2011) South-eastern Identify catch composition and Gillnets Coastal 17 8 Spatial and Queensland, seasonal differences at different beaches seasonal Australia beaches. partitioning.

Abrantes and Southern Determine differences in stable Stomach content Coastal and 1 Possible diet Barnett (2011) Tasmania, isotope composition between sites and stable offshore and habitat Australia to assess connectivity between isotope analyses partitioning. areas and to use SIA, SCA and and electronic tracking data to assess intra- tracking. population differences in diet and movement patterns.

Bethea et al. (2011) Florida, USA Quantify diet and feeding ecology Stomach content Bays 2 Dietary to understand resource partitioning. and stable partitioning isotope analyses.

Powter et al. (2010) South-eastern Determine if there is sex-based and Stomach content Coastal 1 Dietary Australia or ontogenetic differences in diet, analysis temperate partitioning dentition and head morphology. waters

Vaudo and Shark Bay, Spatial and temporal variation in Visual surveys Marine bay 1 10 No clear Heithaus (2009) Western abundance and species composition and capture evidence of Australia and examine size distribution and methods. resource macro-habitat preference among partitioning. common species.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Whitty et al. (2009) Fitzroy River, Test the hypothesis that juveniles Acoustic River 1 Spatial Australia are able to move through shallow tracking partitioning runs to enable upstream migration and that morphological differences in size and age class will mean they will be ontogenetic differences in habitat use.

Saidi et al. (2009) Tunisa Describe food composition in Stomach content Gulf 1 Ontogenetic relation to predator size and season, and stable food determine prey diversity and isotope analysis. partitioning compare with studies from different locations.

DeAngelis et al. US Virgin Assess species diversity and Bottom longline 5 1 Habitat (2008) Islands relative abundance, determine if and hand-gear partitioning area is a nursery habitat and assess sampling. and temporal if habitat partitioning is occurring. partitioning of space use.

Marshall et al. South-eastern To determine if inter-specific Stomach content Demersal 2 Dietary (2008) Queensland, dietary differences occurred analysis partitioning but Australia through ontogeny and if dietary other factors resources partitioning occurred could also play between species. a role.

Taylor and Bennett Moreton Bay, Study trophic relationship between Stomach content Bay 1 cephalopods Specialisation (2008) South-eastern the Australian weasel shark and analysis on cephalopods Queensland, cephalopods. to reduce Australia competition with other sharks.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Treloar et al. South-eastern Comparative feeding ecology and Stomach content Continental 6 Dietary (2007) Australia trophic levels. analysis shelf and partitioning slope among both continental shelf and slope species.

Wiley and Florida, USA Document co-occurring species of Longline, rod Gulf and 4 Habitat Simpfendorfer elasmobranchs, examine the and reel and estuaries partitioning (2007) environmental and habitat factors gillnet. due to that influence occurrence and environmental distribution as well as examine characteristics. movement.

Navia et al. (2007) Colombia Quantify and compare diet and Stomach content Coastal and 1 4 Possible trophic interactions. analysis oceanic temporal and habitat partitioning among rays.

Papastamatiou et Hawaii Quantify and compare diets and Shark control Marine 4 Spatial al. (2006) geographical depth distributions to program data separation determine resource partitioning and and Stomach possibly based or competition. content analysis. on competition.

Pikitch et al. (2005) Belize Assessment of species diversity, use Longline Reef and 10 3 Habitat and of reef system by early life stages, sampling, atoll depth species and age specific patterns of gillnet, seine, partitioning. abundance in reef macro habitats drumline and and demographic population opportunistic structure of early life market surveys. characteristics.

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited

Simpfendorfer et Florida, USA Examine distribution and Longline Estuaries 1 Size based al. (2005) investigate habitat partitioning and surveys habitat the environmental parameters that partitioning may influence distribution. influenced by salinity.

White and Potter Western Test hypotheses that the area is Gillnets Inshore bay 5 4 12 teleosts Spatial and (2004) Australia used as a nursery area, use of area dietary habitat is partitioned, species have partitioning. more affinity to vegetated regions then non-vegetated areas and composition of species that use the area change over time.

White et al. (2004) Western Test hypotheses that food Stomach content Inshore bay 3 1 Dietary Australia partitioning is occurring, different analysis area partitioning

feeding habits and morphology will that could also facilitate differences in diets, be related to species occurring mostly over habitat. seagrass will have different diet, diets changes with body size, main prey will be most abundant crustaceans and prey composition will change throughout the year.

Bethea et al. (2004) Apalachichola Describe and quantify diet to Stomach content Inshore bay 4 Spatial and Bay, Florida, calculate diet overlap, habitat analysis and area temporal USA overlap and investigate prey size- catch data. partitioning. predator size relationship.

Brickle et al. Falkland Islands Examine ontogenetic shifts in diet Stomach content Demersal 3 Dietary (2003) and dietary overlap among species. analysis. partitioning

Study Location Objective Method Habitat Shark Ray Other Outcome: species species species Type of partitioning exhibited and depth partitioning

Platell and Potter Western Compare dietary compositions to Stomach content Temperate 4 14 teleosts Dietary and (2001) Australia analyse inter specific differences to analysis coastal depth assess potential resource partitioning petitioning, also to analyse dietary among species. differences in relations to ontogeny, phylogeny, mouth morphology and feeding behaviour.

Platell et al. (1998) Southwestern Determine if species are partitioned Stomach content Temperate 4 Combination of Australia by habitat, type of food and if there analysis coastal diet and habitat is interspecific competition. partitioning.

2.3 Development of methodologies to study resource partitioning

Ecology is as an empirical study (Krebs, 1989) and multivariate in nature, with many facets contributing to defining the ecological niche of a species (Munroe et al., 2014). To understand the resource use patterns of elasmobranchs across dietary, habitat, and temporal planes, a number of methods have been used either separately or in combination.

Dietary partitioning has been the most widely studied form of resource partitioning among elasmobranchs. Earlier work that addressed the diets of co-occurring elasmobranchs used stomach content analysis (McEachran et al., 1976, Platell et al., 1998). This has since become a common feature of dietary resource partitioning studies. Stomachs are extracted and dissected to identify prey items to the lowest possible taxon and comparisons are made either within members’ of one species or among species to assess the level of diet overlap from which competition is inferred (Wetherbee et al., 2012). However, stomach content analysis does have its drawbacks, the method has been criticised for; providing only a narrow view of what is fed on at the point of sampling and not a broader view of diet over time (Hussey et al., 2012, MacNeil et al., 2005), the predominance of indigestible hard parts (for example from cephalopods beaks and crustacean exoskeletons) that may overestimate the prevalence of certain prey in the diet (Kim et al., 2012), occurrence of empty stomachs particularly in sharks resulting from regurgitation during capture (Cortés, 1997), as well as the need for large sample sizes to adequately characterise diet which means the mortality of large number of individuals. However, some studies that carried out stomach content analysis were able to access large sample sizes by obtaining samples from the bycatch of various fisheries (Rogers et al., 2012, Barbini and Lucifora, 2012, Bornatowski et al., 2014b), or used gastric lavage

(Frisch et al., 2016).

More recently stable isotope analysis has been used in the study of elasmobranchs ecology and resource partitioning. The method has been hailed as a low cost and non-lethal option compared to the acquisition and examination of actual stomachs from dead animals

(Shiffman et al., 2012). In addition, stable isotope analysis provides an indication about longer term trophic ecology compared to stomach content analysis. Essentially stable isotope analysis is based on the process of examining the levels of carbon and nitrogen isotopes in the tissues of consumers that can be related directly to the isotopes of their prey (Kinney et al.,

2011). One of the key strengths of stable isotope analysis is its capability to investigate both dietary and habitat use patterns (Hussey et al., 2012). Over the last 10 years the use of this method has steadily increased, however issues of caution that may affect stable isotope values and consequently the interpretation of results are also prevalent. For instance isotopic values may be influenced by the presence of urea and lipid in samples (Gallagher et al., 2017), changes or fluctuations in the isotopic signatures of base prey in space and over time may also cause variation in carbon and nitrogen isotopic values in consumers (Shiffman et al.,

2019). Though stable isotope analysis provides an indication of average resource use trends, the interpretation of stable isotopes values in isolation from other trophic data can lead to erroneous conclusions, and interpretations from stable isotope analysis should not be extrapolated to the wider ecosystem without complementary stomach content and movement data which provide empirical evidence that can validate or enhance findings (Gallagher et al.,

2017, Shiffman et al., 2019, Abrantes and Barnett, 2011).

Capture methods are often used to study the use of resources in time and space. Fishery independent surveys have been used to capture species at various sites and depths, while some studies have used data from shark control programs (Papastamatiou et al., 2006, Taylor

22 et al., 2011). Tag recapture was also used as a means to investigate habitat use (Wiley and

Simpfendorfer, 2007). The development of electronic devices (e.g. acoustic, data storage and satellite tags) to monitor the movement of individual animals has become widely used to study aquatic animals and provides more fine scale data on the habitat use of some elasmobranch (Donaldson et al., 2014). The majority of studies that use acoustic tracking have been done on sharks while the habitat use patterns of rays have been largely inferred from stable isotope analysis and catch information (Table 2-1).

The need for an integrated approach to studying elasmobranch resource use has been recognised in earlier work (White et al., 2004). Various studies have used complementary data sets using a combination of methods such as both stable isotope and stomach content analysis to determine trophic ecology (Bethea et al., 2011, Vaudo and Heithaus, 2011) or the inclusion of catch data (Gallagher et al., 2017, Papastamatiou et al., 2006) and spatial data to determine both dietary and spatial partitioning (Abrantes and Barnett, 2011). Complementary published stomach content information has also been valuable to validate the results of stable isotope analysis (Matich et al., 2017a). The development of methodologies that can provide a broader assessment of resource use patterns and the improved ability to track the movement of species over time and space has provided researchers with a greater ability to understand and monitor the ecology of elasmobranchs which when coupled with life history and population information is valuable for conservation and management efforts.

2.4 Dietary partitioning

Dietary partitioning is likely to occur among elasmobranchs where species co-exist in a given area (Wetherbee et al., 2012). The majority of studies on elasmobranch resource partitioning have focused on this aspect. Dietary partitioning has been detected among sharks (Bethea et al., 2004, Hussey et al., 2012, Rogers et al., 2012), skates (Treloar et al., 2007, Barbini and

Lucifora, 2012), stingrays (Bornatowski et al., 2014b, Pardo et al., 2015, Platell et al., 1998,

Platell and Potter, 2001) and species assemblages of sharks and rays (Dale et al., 2011,

Kiszka et al., 2014, White et al., 2004, Sommerville et al., 2011, Heithaus et al., 2013) as well as in feeding interactions with other non-elasmobranch species (Kinney et al., 2011, Platell and Potter, 2001).

Interspecific dietary differences occur at varying levels. The same range of prey may be consumed but relative proportions among species differs (Platell et al., 1998). White et al.

(2004) found that three shark species fed predominantly on teleosts, however the species composition of prey was different for each shark species. In other instances there was more specialisation in diets where the prey item constituting the largest portion of each diet was different (Bornatowski et al., 2014b, Sommerville et al., 2011). For example dietary patterns of two sympatric rays were found to be markedly different feeding on benthic crustaceans and polychaetes, respectively (Yick et al., 2011). Furthermore similar co-existing species have diversified their diets to the point where they feed at different trophic levels (Amariles et al., 2017, Raoult et al., 2015).

Ontogenetic food partitioning or the change in diets with growth and maturity among members of the same species has also been widely detected in elasmobranchs (Barbini and

Lucifora, 2012, Sommerville et al., 2011). Morphological traits such as dentition and gape size develop with growth and enable the capture of larger prey (Powter et al., 2010) and can facilitate adults and juveniles to feed at different trophic levels to partition food resources

(Amariles et al., 2017). For instance, adults and juveniles of the Australian weasel shark

Hemigaleus australiensis feed on different prey (Taylor and Bennett, 2008). Prey release by adult members of population when diet changes occur may ensure that there is food supply for young immature individuals and therefore increase the likelihood of survival (Ebert,

2002). However, dietary partitioning between sexes of the same species is rare among elasmobranchs, as Kiszka et al. (2014) was the only study that detected potential intraspecific niche partitioning among sexes.

Prey abundance may be a major contributing factor to the presence of competition for food between co-occurring species (Bornatowski et al., 2014b). However, few studies have directly measured prey abundance and diversity (Pardo et al., 2015, Gutteridge et al., 2011).

Competition is mainly inferred from dietary overlap where sharks and rays with more specialised diets that have little overlap are presumed to be partitioning resources based on the limitation of prey. Conversely, a high dietary overlap can mean that prey is not limited

(Heithaus et al., 2013) or that there is direct competition for prey. The availability of prey may also vary temporally and spatially. Curnick et al. (2019) found dietary partitioning between two shark species was only seasonal due to the influx of prey at certain times of the year while differences in diurnal and nocturnal feeding patterns can partition food among co- occurring species (Varghese et al., 2014). Further information on prey abundance, diversity and distribution will improve the current understanding of elasmobranch feeding ecology.

2.5 Spatial and temporal partitioning

In studies where attempts to investigate dietary partitioning showed a lack of distinction in diets, inferences are made to spatial or temporal partitioning as the likely cause of co- existence (O'Shea et al., 2013, Bethea et al., 2004, Tillett et al., 2014, Navia et al., 2007).

Though stomach content and stable isotope analysis have both provided indirect indication of potential spatial and temporal partitioning (Kiszka et al., 2014, Varghese et al., 2014).

More empirical evidence of elasmobranch habitat use patterns from catch records and movement data suggest that substrate type, competition, abiotic factors, and asynchronous seasonal behaviour are factors that influence habitat partitioning. Elasmobranchs, particularly juveniles, have been observed to have an affinity to vegetated areas compared to unvegetated zones possibly due to the provision of nutrition and shelter from predators (White and Potter,

2004, DeAngelis et al., 2008). Behavioural inter-specific avoidance of competition was exhibited in sharks (Papastamatiou et al., 2006, Bangley and Rulifson, 2017, Heupel et al.,

2019). Competition avoidance can also lead to asynchronous seasonal behaviour. For example one species of shark was found to remain in cooler waters when all other co- occurring species had migrated to warmer regions (Taylor and Bennett, 2013). Physiological tolerance to salinity and temperature have also determined distribution of sharks in nursery areas (Simpfendorfer et al., 2005, Bangley and Rulifson, 2017). Additionally a study on juvenile sawfish found that habitat was partitioned by different aged cohorts with regard to temperature and light regimes due to lunar phases (Whitty et al., 2009).

2.6 Interrelationship between resource use axes

Dietary and habitat partitioning may occur together to facilitate co-existence. For example four species of rays occurring across a 200 km area in Western Australia partitioned both food and space (Platell et al., 1998). White et al. (2004) also documented that both food and habitat was partitioned among an array of 14 species of elasmobranchs in Shark Bay,

Australia. A study conducted on five species of sympatric stingrays found that only one species had a specialised diet implying food partition while the other four species were likely to employ habitat partitioning as a survival strategy that reduces competition (O'Shea et al.,

2013).

The inter-relationship between habitat use and diet is complex and requires further investigation as studies are often focused on a single species or location (Shiffman et al.,

2019). Pardo et al. (2015) suggested that large-scale sampling may overlook microhabitat partitioning which could cause the dietary differences observed among species. A similar view was provided by Marshall et al. (2008) in explaining the reason behind the significant dietary differences of sympatric urolophid rays. Bornatowski et al. (2014b) found evidence of dietary partitioning but suggested that further study of spatial and temporal distribution should also be conducted. Conversely, the movement patterns of elasmobranchs may not always reflect resource use and wider data sets including dietary information is needed to verify observations (Heithaus et al., 2013). This emphasises the need for integrated studies across all resource axes (White et al., 2004).

Recent studies show more evidence of the dynamics of elasmobranch resource partitioning.

For example the feeding habits of two juvenile shark species were compared in habitats where they both existed and where only one species was present; dietary partitioning only occurred where both species were present and competing for resources (Matich et al., 2017b).

Furthermore, habitat type can play a significant role, an assemblage of elasmobranchs present in distinct habitats that were in close proximity to each other partitioned resources only in one habitat type (Shiffman et al., 2019). Therefore fine-scale assessment of resource use is important to identify variability and generalisations should be avoided (Shiffman et al.,

2019).

2.7 Future work

The segregation of resource use occurs according size, sex, morphology, and behaviour of a species or by social interactions such as competition, and also is influenced by abiotic environmental elements. The current body of work has been built upon characterisations of food or space use by elasmobranchs to document more complex and changing patterns in interactions between species. Despite these advancements there are knowledge gaps that remain in understanding elasmobranch ecology from the standpoint of resource partitioning.

These include: (1) a lack of studies conducted in tropical areas that support a large diversity of elasmobranchs as well as other fish species and marine taxa, (2) literature on spatial and temporal use of resources is more limited than dietary information, though stable isotope analysis improves the understanding of habitat use, more empirical movement data is needed to accurately represent the space use of many species, particularly rays, (3) the extent to which abiotic factors influence species distribution needs more attention, particularly in light of other driving factors such as climate change and the impact of development in coastal

28 habitats, (4) the resource use patterns of elasmobranchs in coastal and nearshore areas is more common in literature, however further work in pelagic and deep-water species is needed to develop a better understanding of species ecology in these habitats. Finally, the high variability of ecosystem function, habitats, and species interactions requires more integrated assessments of species ecology utilising multiple complementary data sets to arrive at a more accurate representation of resource use patterns and the factors that play a critical role in influencing the outcomes.

Chapter 3 Dietary overlap of carcharhinid sharks in the Gulf of Papua

3.1 Introduction

Fisheries are a major contributor to the decline of shark populations (Dulvy et al., 2014) that function mainly as top and middle order predators in marine ecosystems (Heupel et al.,

2014). Concern for the survival of these populations has also highlighted that the flow on effects of low predator abundance on the ecosystem remain largely unknown, partly due to the paucity of ecological information for specific regions (Ferretti et al., 2010). Therefore, establishing an understanding of the ecology of species, and their contributions to ecosystem processes (Bornatowski et al., 2014a), is a crucial element in predicting the outcomes of population declines and potential species loss. Assessing the ecosystem impacts of fisheries in order to set appropriate management and conservation guidelines requires information from both target and non-target (bycatch) species (Pikitch et al., 2004).

Characterising the dietary traits of sharks from stomach content analysis provides empirical evidence of the trophic linkages in the food chain (Cortes, 1999). This information can be incorporated into ecosystem models to aid fisheries management endeavours (Rogers et al.,

2012). Furthermore feeding patterns indicate diet specialisation, which helps understand the vulnerability of predators and the breadth of ecosystem impacts from their decline.

Specialised feeders have a narrow range of prey and may be more vulnerable to perturbations that may directly impact food availability while generalist feeders may be more resilient to environmental changes (Munroe et al., 2014, Simpfendorfer et al., 2001). The level of diet overlap among similar sympatric species is an indirect measure of competition among species

30 when food resources are limited, and also provides an indication of potential resource partitioning among species (Wetherbee et al., 2012). Dietary studies have shown that potential competition for food can influence the differential distribution of similar shark species (Papastamatiou et al., 2006) while evidence of resource partitioning as a possible means to alleviate competition is common in sharks (Wetherbee et al., 2012). Dietary investigations thus provide a preliminary view of complex and dynamic ecological interactions that require integrated datasets (White et al., 2004), and monitoring diets over time can gain an understanding of the ecology of a species and its role in the overall function of the ecosystem.

Small-bodied coastal sharks are generally considered to be meso-predators that connect the lower and top trophic levels of the food chain (Heupel et al., 2014) and are also common in fisheries bycatch (Stobutzki et al., 2002). The Australian blackspot shark Carcharhinus coatesi, the milk shark Rhizoprionodon acutus and the Australian sharpnose shark

Rhizoprionodon taylori are small-bodied coastal sharks that are frequently caught as bycatch in the Gulf of Papua Prawn Fishery (GoPPF) in Papua New Guinea (White et al., 2019). The life histories of C. coatesi and R. taylori indicate that the populations of each species may be impacted differently by the fishery based on growth and biological productivity (Baje et al.,

2018, Baje et al., 2019). However, the ecology of these sympatric sharks has not been investigated in the Gulf of Papua, and their ecological roles are not well understood. Using samples caught in the fishery this study aimed to characterise the diets of C. coatesi, R. acutus and R. taylori and estimate the level of dietary overlap to assess if competition and partitioning of food resources occurs among these species in the Gulf of Papua. We hypothesise that the diets of these species will be different based on the occurrence of resource partitioning.

3.2 Methods

3.2.1 Study site

The Gulf of Papua, situated on the south coast of Papua New Guinea (Fig 3.1), is a region comprised of extensive mangrove and estuarine areas with high riverine input. Waterways from high altitude areas of PNG drain into the Gulf forming several major river systems, the largest of which is the Fly River in the West. North Eastward of the Fly River are the Kikori and Purari rivers along with several other systems. These areas provide major nursery grounds for penaeid prawn species which eventually recruit into the Gulf of Papua prawn trawl fishery (Evans et al., 1997). The region experiences two main seasons: the North-West monsoon from November to March each year and the South-East monsoon winds that occur from April to October

(Moore and MacFarlane, 1984).

Figure 3-1: The Gulf of Papua situated in the south of Papua New Guinea.

3.2.2 Sampling and sample preservation

Fishery observers were deployed on seven prawn trawl fishing trips between June 2014 and

August 2015 to collect shark bycatch samples. Samples were kept whole and frozen on board.

In a laboratory, sharks were thawed, total length (TL) measured to the nearest ± 1 cm, sex recorded and stomachs excised. The level of fullness was estimated and graded as: an empty stomach = 0, 25% full =1, 50% full = 2, 75% full = 3 and 100% full = 4 was recorded for each sample, contents from each stomach were removed, fixed in 10% formalin and transferred to

70% ethanol for preservation. Each set of stomach contents were weighed and examined to identify the number and type(s) of prey to the lowest possible taxa. The level of digestion was classified using a grading system from 1–5 based on the amount of body tissue of the prey remaining as follows: 100–80% was classified as stage 1, 80–61% was classified as stage 2,

60–41 % stage 3, 40–21% stage 4 and 20–1% stage 5 (Simpfendorfer 1993). In order to detect if the sample size was sufficient to adequately describe diets, a cumulative prey curve was produced using the specaccum function of the ‘vegan’ package (Oksanen et al., 2013) in R (R

Core Team, 2015).

3.2.3 Dietary indices

To assess the importance of each prey item in the diet of the three shark species the percent frequency of occurrence (% FO) and the percent by number (%N) were calculated. The former is the number of times a prey category is present in one or more stomachs expressed as a percentage of the total number of stomachs containing food while the latter is the number of each prey category found in each stomach expressed as a proportion of the total number of prey for all stomachs of a particular species (Hyslop, 1980). The state of digestion and mastication in most of the samples meant that prey items could not be adequately identified and separated therefore volumetric and gravimetric methods were not carried out.

3.2.4 Dietary overlap

Dietary overlap, which is a measure of the level of similarity in the diets between shark species, was measured using the simplified Morisita index (Krebs, 1989):

푛 2Σ푖 푝푖푗 푝푖푘 퐶퐻 = 푛 2 푛 2 Σ푖 푝푖푗 + Σ푖 푝푖푘 where CH = Simplified Morisita index of overlap between two species with values ranging from 0 (no overlap) to 1 (complete overlap).

pij = the proportion prey in species i that is of the total prey categories used by species j

34 pik = Proportion prey i is of the total prey categories used by species k. n = total number of prey categories.

3.2.5 Multivariate analysis

Samples were initially randomised and pooled within each species to minimise the large number of zeros and improve the effectiveness of the analyses (Sommerville et al., 2011).

The resulting new samples comprised stomach contents from 4 or 5 individuals of the same species randomly pooled together. The percentage by number (%N) was calculated for each prey item in each sample and entered into Primer-E (Plymouth Routines in Multivariate

Ecological Research) version 7.0. 13 (Clarke and Gorley, 2015). Prior to further analysis the data were subject to square-root transformation followed by creation of a Bray Curtis resemblance matrix. To test for differences in dietary composition among sex, species and season Analysis of Similarities (ANOSIM) was conducted. Similarities of Percentages

(SIMPER) was also used to identify the components that typified the diets of each shark species. Non-metric Multidimensional Scaling (nMDS) ordination was used to produce plots to visualise the dietary composition based on species, size and season. To test for the multivariate variability in the diet of each species Multivariate Dispersion (MVDISP) was conducted. To assess if the diets of the three shark species undergo changes with respect to growth, samples were grouped into 10 cm size classes and randomly pooled into groups of 4 or 5 samples in the data preparation stage. Column graphs were constructed to demonstrate any change in the composition of diets with respect to increasing size and an nMDS ordination plot was also used to visualise the level of similarity in diets between different size classes.

3.3 Results

3.3.1 Size ranges and sample size

Total lengths recorded were similar among species and ranged from 31–76 cm TL for R. taylori; 31–84 cm TL for R. acutus; and 35–79 cm TL for C. coatesi (Fig 3-2). A total of 177 stomachs were sampled of R. taylori, 83 of R. acutus, and 122 of C. coatesi. The cumulative prey curve for all three species did not appear to reach asymptote, indicating a larger sample size would be required to fully characterise the diets (Fig 3-3). The number of stomachs containing prey was high with few empty stomachs encountered for each species (Table 3-1).

Figure 3-2: Length frequency of Carcharhinus coatesi, Rhizopriondon acutus and

Rhizoprionodon taylori caught in the Gulf of Papua prawn trawl fishery and used for stomach content analysis

Figure 3-3: Cumulative prey curves for Carcharhinus coatesi (blue) Rhizoprionodon acutus

(green) and Rhizoprionodon taylori (red) from the Gulf of Papua.

3.3.2 Main prey types and proportions in diet

Teleosts, crustaceans and molluscs were observed as the main prey groups, with sixteen teleost families, three crustacean families and two families of molluscs identified. The proportion of individual %FO and %N of each teleost family was low, not exceeding 5% owing to mastication and the process of digestion that resulted in only a small number of individual fishes being identified. Of the 16 families of teleosts observed only 3 families:

Haemulidae, Engraulidae and Trichiuridae appeared in the diet of all three shark species.

Other families were only shared between two of the species, for example Leiognathidae was

38 only present in the stomach contents of R. acutus and C. coatesi. However, distinctively the families Pegasidae, Fistulariidae and the eel families Muraenesocidae and Ophichthidae, were only present in the diet of C. coatesi. The proportion of unidentified teleosts was high for all species (Table 3-1).

The presence of crustaceans in the diet, %FO and %N of Penaeidae was high for all species, but particularly prevalent in the diet of C. coatesi (54.5 %FO). Stomatopoda were also more common in the diet of C. coatesi (23.14%FO and 7.51%N) compared to R. taylori

(11.61%FO and 5.36 N) while being absent in the diet of R. acutus. Similarly, with respect to crabs there was a higher %FO and %N in the diet of C. coatesi (7.44% FO and 2.15 % N) compared to R. taylori (1.29 % FO and 0.59% N) and R. acutus (1.33% FO and 0.5% N).

Molluscs played a lesser role in the diet of the all three species, R. acutus (2.67% FO and

0.99% N) consumed fewer cephalopods than R. taylori (7.74 % FO and 5.06% N) and C. coatesi (5.78 % FO and 6.22 %N) while Gastropoda were only found in the stomach contents of R. taylori (Table 3-1).

Table 3-1: Percent frequency of occurrence (% FO) and percent by number (% N) of prey categories found in the stomachs of Rhizoprionodon taylori, Rhizoprionodon acutus and

Carcharhinus coatesi in the Gulf of Papua

Prey categories R. taylori R. acutus C. coatesi

%FO %N %FO %N %FO %N

Teleostei

Sciaenidae 3.1 0.6 2.7 2.0 - -

Labridae 1.6 0.3 - - - -

Mullidae 1.6 0.3 - - - -

Haemulidae 3.1 0.89 1.3 2.0 1.65 0.43

Engraulidae 1.55 0.3 2.67 0.99 0.83 0.215

Nemipteridae 1.29 0.6 - - - -

Gobiidae 1.94 0.89 - - 0.83 0.22

Synodontidae 1.29 0.6 1.33 0.5 - -

Terapontidae 1.29 1.2 2.67 1.0 - -

Trichiuridae 0.65 0.3 2.67 0.5 0.83 0.22

Carangidae 0.65 0.3 - - - 0.22

Leiognathidae - - 4.00 1.99 4.13 1.07

Pegasidae - - - - 0.83 0.22

Fistulariidae - - - - 0.83 0.22

Muraenesocidae - - - - 1.65 0.43

Ophichthidae - - - - 0.83 0.22

Unidentified eel - - - - 0.83 0.24

Unidentified teleost 56.77 45.24 77.33 77.11 54.5 44.42

Crustacea

Penaeidae 36.77 27.68 25.33 10.95 51.24 33.05

Stomatopoda 11.61 5.36 - - 23.14 7.51

Crab 1.29 0.59 1.33 0.5 7.44 2.15 unidentified crustacean 6.45 3.57 1.33 0.5 - -

Mollusca

Cephalopoda 7.74 5.06 2.67 0.99 5.78 6.22

Gastropoda 1.29 0.89 - - - -

Other unidentified 20.15 4.17 1.33 0.5 7.44 2.15

No. of stomachs analysed 177 83 128

No. of stomachs with food 155 75 121

No. of empty stomachs 22 8 7

3.3.3 Diet Overlap

The Morisita Index of Similarity calculated for each pair of species showed high overlap for all species. The highest overlap was between R. taylori and C. coatesi (CH = 0.99) with less overlap between the diets of R. taylori and R. acutus (CH = 0.85) and R. acutus and C. coatesi

(CH = 0.82).

3.3.4 Multivariate analyses

3.3.4.1 Intraspecific dietary comparison

Dietary data for males and females were pooled for subsequent analysis as there was no significant difference between sexes (P = 0.4, R= 0.039). A one-way ANOSIM indicated a significant difference among the diets of R. taylori, R. acutus and C. coatesi (P= 0.1, R=

0.181). Similarities of percentages (SIMPER) showed that the main groups that typified the diets of R. taylori and C. coatesi were unidentified teleosts and penaeid prawns, while unidentified teleosts typified the diet of R. acutus. The pairwise tests between species showed a significant difference in dietary compositions of R. acutus and R. taylori (P = 0.2, R =

0.243) and R. actus and C. coatesi (P = 0.1, R = 0.479). However, there was no significant difference in dietary compositions between R. taylori and C. coatesi (P > 0.05, R = 0.28). The multivariate dispersion (MVDISP) analysis showed that R. taylori had the highest dispersion of 1.17, followed by C. coatesi with 0.81 and R. actus with 0.53. The nMDS ordination plot of the dietary compositions of the three shark species showed that R. taylori has a broad diet that overlaps with C. coatesi and also with R. acutus. Samples of R. acutus appeared in the bottom left of the plot and did not overlap with C. coatesi (Fig 3-4).

Figure 3-4: non-Metric multidimensional scaling ordination of dietary composition by number (%N) of Rhizoprionodon taylori, Rhizoprionodon acutus and Carcharhinus coatesi in the Gulf of Papua.

3.3.4.2 Dietary comparison by season

A one-way ANOSIM testing between North-West Monsoon and South East Monsoon periods did not detect a significant result (P >0.05, R = -0.002) indicating there was no difference in the diets of all three species between seasons. The nMDS ordination of diets sampled in different seasons showed that the majority of South East Monsoon samples overlapped with

North West Monsoon indicating similarity (Fig 3-5).

Figure 3-5: non-Metric multidimensional scaling ordination of dietary composition by number (%N) according to North-west monsoon and South-east monsoon periods that occur in the Gulf of Papua.

3.3.4.3 Dietary composition among size classes

Comparison of diet composition among size classes for each species showed that R. taylori has a fairly consistent diet with respect to proportions of different dietary components.

Cephalopods were not consumed by the smallest size class and there may be a reduction in the consumption of penaeid prawns in the largest sizes class with a possible increase in the consumption of teleosts. Rhizoprionodon acutus consumes large proportions of teleosts in all size classes and may consume less crustaceans and cephalopods with increasing size.

Carcharhinus coatesi had a marked decrease in teleost consumption with increasing size accompanied by an increase in the consumption of crustaceans particularly penaeid prawns

(Fig 3-6).

Figure 3-6: Composition of the diets of Carcharhinus coatesi, Rhizoprionodon acutus and

Rhizoprionodon taylori according to different size classes.

The nMDS ordination plot of size classes showed a more pronounced difference among species, with the vast majority of the samples of R. taylori and C. coatesi clustering on the left of the plot and well separated from the majority of R. actus samples on the right. Among

R. taylori and C. coatesi cluster the 41–50 cm size class and the 51–60 cm size class were more dispersed, while distinctively the largest size class (71–80 cm) of C. coatesi appeared to the left of the plot and the largest size class (61–70 cm) of R. taylori was situated away from the main group of samples indicating less similarity. For R. acutus the 51–60 cm and 61–70

46 cm size classes were not available due to a lack of samples. However, there was some separation among the 41–50 cm, 71–80 cm and the 31–40 cm size classes (Fig 3-7).

Figure 3-7: non-Metric multidimensional scaling ordination of dietary composition by number (%N) of size classes of Rhizoprionodon taylori (1), Rhizoprionodon acutus (2) and

Carcharhinus coatesi (3) in the Gulf of Papua.

3.4 Discussion

Many shark species are considered to be generalist feeders (Munroe et al., 2014) and have been observed to feed in a density-dependent manner (Salini et al., 1992). The small-bodied carcharhinids studied here feed at similar trophic levels (Cortes, 1999), therefore where they co-occur competition for food resources can arise if prey are limited. This study shows that teleosts, crustaceans and molluscs make up the majority of prey in the diets of C. coatesi, R. acutus and R. taylori. There are noticeable differences in the diet of these species that may facilitate co-occurrence. Teleosts and greater proportions of crustaceans were found in the diet of R. taylori and C. coatesi while the diet of R. actus consisted predominantly of teleost with other prey categories being much less important. Stevens and McLoughlin (1991) found similar predominant prey types for all three species in northern Australia, however the relative amounts of prey differed from this study. Furthermore the findings of this study align with the classification of R. taylori as dietary generalist due to its broad diet breath (Munroe et al., 2015b) and additionally a predominance of teleosts in the diet of R. acutus (Ba et al.,

2013).

Resource partitioning in elasmobranchs is a common occurrence (Wetherbee et al., 2012) and can occur in different ways. Co-occurring similar species that have high spatial overlap may partition resources by consuming different proportions of prey available to them (Platell et al., 1998, White et al., 2004). Significant dietary differences and low level of similarity indicate that R. acutus partitions food resources with C. coatesi by feeding predominantly on teleosts and reducing consumption of crustaceans, molluscs and other groups. The greater reliance of R. acutus on teleosts has been noted in previous studies in Australia and Africa

(White et al., 2004, Stevens and McLoughlin, 1991, Ba et al., 2013). Rhizoprionodon acutus also occupies a broader depth range (White et al 2017) which may provide a larger foraging area. There was high dietary overlap and no clear partitioning between C. coatesi and R. taylori however, noticeably C. coatesi consumed a few prey groups that were not observed in the R. taylori diet, further sampling would be needed to investigate these differences as the cumulative prey curves for all species indicated that the sample size was not sufficient to fully describe diets.

Apart from interacting with a wide range of bycatch species, including elasmobranchs, trawl fisheries also contribute to the disturbance of the benthic environment and provide or expose unnatural sources of food (Dayton et al., 1995) for opportunistic feeders, which can impact community structure (Kaiser and Spencer, 1994). Sharks have been observed to scavenge on discarded trawl catch (Hill and Wassenberg, 1990). The typical diet of inshore sharks is mainly made up of teleosts, crustaceans and molluscs (White et al., 2017b) captured in trawl grounds (Stobutzki et al., 2002). The comparison of diets from fishery independent sampling may be an option for future work to explore the impact of trawling on the diets of common bycatch species which may investigate possible changes to the benthic community structure.

Volumetric and gravimetric (bulk) descriptions of diets have been consistently included with other measures to produce compound indices and have been the preferred measure on which to conduct multivariate analysis. However, practically assessing stomach contents to achieve bulk measures of diets is associated with the difficulty of sorting through masticated and partially digested prey items that are separated into many pieces or loose tissue which makes it impossible to know which prey item they belong to or if they are part of a separate prey

49 item altogether. Thus, the inclusion of bulk dietary measures introduces inherent errors linked to the difficulty in identifying and quantifying prey items (Baker et al., 2014). The absence of a bulk measure of the diet meant that a compound index (e.g. percent index of relative importance) was not calculated for this study. Compound indices have been recommended as a standard practice (Cortés, 1997, Brown et al., 2012, Hyslop, 1980), however, they have been found to have little significance, as opposed to considering separate dietary measures individually (Baker et al., 2014), particularly for demersal species (Macdonald and Green,

1983).

The patterns of predation in tropical inshore areas are driven by habitat type and abiotic factors which influence the species composition of predator and prey species (Salini et al.,

1998). The Gulf of Papua presents a unique system within PNG that historically hosts a wide array of marine resources (Pernetta and Hill, 1981). The biodiversity and ecological dynamics of this region remain to be fully explored. The stomach content analyses used in this study provided a preliminary understanding of the diet of small-bodied carcharhinids though a larger sample size will be needed to fully characterise the diets of these species because of the high prey diversity and high proportion of unidentifiable stomach contents. This study is also limited in its use of current methodology that could provide a more in-depth assessment of diets through a finer level of prey identification. Recent approaches for dietary analysis have included the use of a combination of different methods alongside stomach content analysis such as stable isotope analysis and molecular techniques to identify prey (Matley et al.,

2018). Future work in the region should consider employing such methods as well as understanding the spatial resource use patterns to draw a clearer picture of the food web and ecosystem use.

Chapter 4 Age, growth and maturity of the Australian sharpnose shark Rhizoprionodon taylori from the Gulf of Papua

4.1 Introduction

A general view on the life history characteristics of sharks assumes slow growth, late maturity, and a low number of offspring resulting in populations that have low intrinsic rates of population growth and are highly vulnerable to overfishing (Stevens et al., 2000, Smith et al., 1998). However, not all shark species share these characteristics. In particular, small- bodied carcharhinids such as the milk shark Rhizoprionodon acutus and the sliteye shark

Loxodon macrorhinus are characterised by relatively rapid growth and early maturity resulting in higher population turnover rates (Gutteridge et al., 2013, Harry et al., 2010). Fast population turnover rates for these species make them potentially more resilient to fishing

(Goldman et al., 2012), although sustainable shark catch is mostly associated with the development of science-based fisheries management (Simpfendorfer and Dulvy, 2017).

The Australian sharpnose shark Rhizoprionodon taylori is a small carcharhinid species known to have one of the fastest growth rates of all shark species (Cortés, 2004,

Simpfendorfer, 1993). Initial studies suggested it grows rapidly in the first year of life, on average increasing to 140% of its length-at-birth, and attains a maximum length of only 67 and 97 cm TL respectively in different locations in Australia (Simpfendorfer, 1993, Taylor et al., 2016). Maturity is reached after only one year with a litter of 1–10 pups produced every year following maturity (Simpfendorfer, 1992, Simpfendorfer, 1993). Rhizoprionodon taylori is also one of the few elasmobranch species that can halt embryonic development (diapause), possibly to facilitate increased litter sizes (Waltrick et al., 2012, Simpfendorfer, 1992).

Occurring only in southern New Guinea and tropical and sub-tropical nearshore waters of

Australia from Carnarvon in Western Australia to Moreton Bay in southern Queensland, it is a locally abundant species often incidentally caught in trawl and gillnet fisheries (Harry et al.,

2011b, Last and Stevens, 2009).

All known biological information about R. taylori has been established from populations in

Australia (Simpfendorfer, 1992, Simpfendorfer, 1993, Simpfendorfer, 1999, Stevens and

McLoughlin, 1991, Simpfendorfer, 1998, Taylor et al., 2016). Recent trawl fisheries data from Papua New Guinea (PNG) confirm that R. taylori is also frequently caught as bycatch in the Gulf of Papua (GoP) (NFA unpublished data). Prawn trawling has occurred in the area since the late 1960’s and bycatch levels can comprise up to 85% of the total catch (Matsuoka and Kan, 1991). However, the effect of trawling on the sustainability of bycatch populations cannot be properly assessed without determining species compositions and locally relevant biological parameters.

Life history traits can differ for populations in separate localities (Lombardi-Carlson et al.,

2003, White, 2007b). The GoP is in close proximity to the northern coast of Australia.

However, R. taylori has been observed to maintain residency in embayments and nearshore habitats, travelling short distances and rarely moving greater than 100 km within 6 months to one year (Munroe et al., 2015a). These limited movements mean that there may be differences in the life history of this species between the GoP and other regions. These differences need to be investigated since variations in size at birth and length-at-maturity could affect fisheries risk assessments and have already been documented between different

52 locations in Australia (Stevens and McLoughlin, 1991, Simpfendorfer, 1992, Taylor et al.,

2016).

Age and growth studies provide essential information for wider population analyses such as stock assessments (Cortés et al., 2012). Growth parameters for R. taylori were determined by

Simpfendorfer (1993) prior to the development and use of multiple growth models within an information theoretic framework, which is now the recommended approach for age and growth studies (Smart et al., 2016b, Goldman et al., 2012). This study used the more contemporary multi-model approach to determine growth and maturity parameters for R. taylori in the GoP. The specific aims were: (1) to determine the age, growth and maturity of

R. taylori; (2) compare life history parameters to previous work to determine if the use of the multiple model approach substantially changed the outcomes; and (3) examine spatial variation in life history of this species. This study also contributes new knowledge from a data poor region that can be used to inform fisheries management and conservation in PNG.

4.2 Materials and methods

4.2.1 Sample collection

This work is a collaboration with the National Fisheries Authority (NFA), the government agency responsible for managing commercial fisheries and implementing fisheries research in

PNG. Fishery observers were stationed on board prawn trawlers and collected sharks that were caught as bycatch and discarded. The sharks collected for this study had already

53 suffered mortality in the process of fishing and no sharks were intentionally sacrificed for the study. All sampling procedures were allowed by the NFA and in line with James Cook

University, Animal Ethics approval A2310 obtained prior to the commencement of the study.

Sampling did not involve endangered or protected species. No further permits were required by relevant authorities.

Commercial trawling in the GoP occurs between Parama Island in the West, just south of the mouth of the Fly River, and the border of the Central and Gulf Provinces in the East (Fig 4-

1). Trawl fishing is permitted all year round throughout the GoP except in a section of the

Gulf between Iokea and Cape Blackwood which is closed to fishing between the 1st of

December and the 31st of March, a measure put in place to protect the growth and survival of prawn recruits (Evans et al., 1997). Samples of R. taylori were collected on commercial vessels from June 2014 to August 2015. Whole samples were kept frozen and brought ashore at the end of each trip for confirmation of identification and processing. In a laboratory samples were defrosted, total length (TL) measured, and sex recorded. For each individual, maturity was also determined using an index modified from (2005a). Reproductive organs were examined and categorised according to the developmental stage of the ovaries and uteri in females, and claspers in males. Females were categorised into one of five stages and males into one of three stages (Table 4-1). A section of the vertebral column from beneath the first dorsal fin was retained and stored frozen for subsequent age determination (Cailliet and

Goldman, 2004, Goldman et al., 2012).

Figure 4-1: The Gulf of Papua is situated along the southern coast of Papua New Guinea. The insert shows the distribution of Rhizoprionodon taylori in Australia.

Female stage Description Binary category

1 Immature Uteri very thin, ovaries small 0 and without yolked eggs.

2 Maturing Uteri slightly becoming 0 enlarged at one end, ovaries becoming larger and small yolked eggs developing.

3 Mature Uteri large along entire 1 length, ovaries containing some large yolked eggs.

4 Pregnant Uteri containing embryos or 1 large eggs.

5 Post-partum Uteri very large but without 1 embryos.

Male stage Description Binary category

NC Not Calcified Clasper very short not 0 extending past the pelvic fin tip.

PC Partially Claspers longer, extending 0 Calcified past the pelvic fin tip, not entirely hard, still flexible.

FC Fully Calcified Claspers long, hard along 1 almost the entire length.

4.2.1 Vertebrae preparation

Vertebrae processing and ageing followed protocols described by Cailliet et al. (2006).

Frozen vertebrae were thawed and any excess tissue was removed using a scalpel. Vertebrae were separated into individual centra and immersed in 4% sodium hypochlorite solution for 3

– 5 minutes to clean remaining soft tissue from the small sized vertebrae. The centra were then rinsed using water and dried in an oven at 60 °C for 24 hours. A single centrum was selected from each individual and mounted on a microscope slide using Crystal bond adhesive (SPI supplies, Pennsylvania, USA). To achieve the desired thickness of < 400 µm the vertebrae was sanded towards the centre of the centrum using 400-1200 grit wet and dry abrasive paper. After one side was complete the centrum was remounted and sanded again on the other side until the desired thickness was achieved (Simpfendorfer, 1993).

4.2.2 Age determination

To estimate the age of each individual, mounted sections of vertebrae were observed using a dissecting microscope. Growth increments appeared as a pair of alternating wide opaque band and a narrow translucent band, referred to as a band pair (2006, Goldman et al., 2012).

The birthmark was identified where there was an obvious change in angle along the corpus calcareum. Subsequent band pairs that spanned from one side of the corpus calcareum to the other side were interpreted to represent annual growth (Cailliet and Goldman, 2004, Goldman et al., 2012). The age of each individual was estimated as the number of band pairs present after the birthmark. The annual deposition of bands for this species has been validated using marginal increment analysis and size frequency data by (1993).

Precision and bias

Visual estimation of age from vertebrae is an approach which may include some level of bias

(Cailliet and Goldman, 2004). To minimise bias two readers estimated ages separately. The first reader conducted an initial read of all vertebrae followed by a second experienced reader. Both readers had no prior knowledge of the sex or size of individuals. Final ages were the result of a consensus process between the readers – where counts were different readers examined the section and agreed on a final age. Where differences could not be resolved, those centra were removed from the analyses. To assess the precision of counts the average percent error (APE) (Beamish and Fournier, 1981), Chang’s coefficient of variation (CV)

(Chang, 1982) and percent agreement (PA ± 1 year) (Cailliet and Goldman, 2004) were used.

Bowker’s test of symmetry was used to estimate bias between readers (Bowker, 1948).

Analyses were carried out using ‘FSA’ package version 0.8.11 in the R program environment version 3.2.2 (R Core Team, 2015).

4.2.3 Partial ages

For a species that reproduces seasonally, and the period of parturition is known, it is possible to assign partial ages and therefore improve age estimation (Smart et al., 2013). The pupping season for R. taylori was observed in January in Queensland (Simpfendorfer, 1993). In this study the largest embryo (22 cm TL) was caught in the month of December, confirming a similar timing in the GoP. Partial ages were calculated by choosing a birth date of 15th of

January and determining the total number of days between this date and the date of capture which was then divided by the number of days in a year. This value was added to the number

58 of full annual band pairs for each individual to give the final age. For example, samples aged at 1 year caught on the 17th of June and 30th of August, respectively, were given partial ages of 1.39 and 1.62 years.

4.2.4 Growth model fitting

The growth of R. taylori was modelled using a multi-model approach. This method incorporated the Akaike Information Criterion (AIC) (Akaike, 1973) which selected the best model fit based on the lowest AIC value (Smart et al., 2016a). Preference for the use of multiple growth models over an a priori approach, using only the von Bertalanffy growth model (VBGM) is standard methodology in elasmobranch growth literature (Smart et al.,

2016b). The multi-model approach is considered to provide better growth estimates as it avoids model mis-specification and biases compared to the use of a single model (Cailliet et al., 2006, Thorson and Simpfendorfer, 2009, Smart et al., 2016b). The lack of small juveniles in the sample, and their likely very rapid growth required a variety of approaches to determine the most suitable growth parameters. Three candidate models were used: VBGM, logistic model, and Gompertz model (Table 4-2). However, because of the limited data from very young individuals three approaches to fitting the models was used: (1) standard three- parameter growth models, (2) versions of the growth models with a fixed length-at-birth

(which ensured that models accounted for the rapid early growth; two-parameter version)

(Harry et al., 2011a), and (3) three-parameter models with four hypothetical aged zero individuals (L0 = 26 cm TL) added to the sample in order to provide a reference point for the model given that aged zero individuals were absent from the sample (Smart et al., 2013).

Separate growth models were constructed for males, females, and combined sexes.

Table 4-2: Equations of the three growth functions used in the multi model approach

Model Growth function

(−푘푡) von Bertalanffy 퐿(푡) = 퐿0 + (퐿∞ − 퐿0)(1 − 푒 )

퐿∞퐿0(푔(log )푡) 퐿(푡) = (푔(log )푡−1) Logistic 퐿∞ + (퐿0 푒 )

(−푔(푔표푚)푡) Gompertz −퐿0 푒 퐿(푡) = 퐿∞ 푒( )

The three-parameter models estimated length-at-birth (L0), asymptotic length (L∞) and the different growth coefficients for each respective model; k indicates the relative growth rate of the VBGM model while g(log) and g(gom) represent alternative sigmoidal growth of the

Gompertz and logistic models (Katsanevakis and Maravelias, 2008). The two-parameter models incorporated a fixed known value for length-at-birth and thus the models only estimated the asymptotic length and the growth coefficients. Umbilical scars were not recorded in this study which meant that a length-at-birth for R. taylori in the GoP was not identified but could be estimated using other data available from the sample as well as published information. In this study the smallest free swimming individuals were 31 cm (TL) and largest embryos were 22 cm (TL) observed in December (a month prior to pupping). The literature estimates of length-at-birth are 25–30 cm (Stevens and McLoughlin, 1991) from northern Australia and 22–26 cm in north eastern Australia (Simpfendorfer, 1993). A possible estimate for the length-at-birth would therefore be 22–30 cm, however in the GoP R. taylori are still embryos at 22 cm and are possibly born at a larger size. The midpoint between 22 and 30 cm (26 cm) was chosen because this value was within the length-at-birth range

60 suggested by both previous studies and was biologically plausible given embryo sizes in the

GoP. Growth models were fit using the ‘nls’ function, multi-model analysis was conducted using the ‘MuMIn’ package version 1.15.6 (Barton, 2016) and bootstrapped confidence intervals were produced using the ‘nlstools’ package version 1.0-2 (Baty et al., 2015) in the R program environment version 3.2.2 (R Core Team, 2015).

As the sample size was less than 200, the AICC, a size adjusted bias correction, was used

(Zhu et al., 2009) :

2푘(푘 + 1) 퐴퐼퐶 = 퐴퐼퐶 + 퐶 푛 − 푘 − 1 where 퐴퐼퐶 = 푛푙표푔(휎2) + 2푘, k is the total number of parameters + 1 for variance σ2 and 푛 is the sample size. The model that has the lowest 퐴퐼퐶퐶 value (퐴퐼퐶min ) was chosen as the best fit for the data. The AIC difference (∆) was calculated for each model (i = 1-3) and used to rank the remaining models as follows:

∆푖= 퐴퐼퐶퐶푖 − 퐴퐼퐶푚푖푛

Models were ranked according to the value of ∆. Values from 0-2 were considered to have the strongest support, less support was given to values between 2-10 and the least support for

∆ values > 10 (Anderson and Burnham, 2002). The AIC weights were calculated by the expression:

∆ − 푖 푒 ( 2 ) 푤푖 = ∆ 푖 3 2 (∑푗=1 푒( ))

To test if there were differences in the growth curves for males and females, a likelihood ratio test was carried out (Kimura, 1980). This was conducted on the model with the best fit based on the AICC results for the sexes combined. The method used to carry out the likelihood ratio test was described by (Haddon, 2001) and incorporated into the R program environment version 3.2.2 (R Core Team, 2015) for this analysis.

4.2.5 Maturity

The maturity stage data was converted to a binary maturity category (immature = 0 or mature

=1) for statistical analyses (Table 4-1). The length-at-maturity was estimated for both males and females using logistic regression (Walker, 2005a) :

−1 푙−푙 − ln(19)( 50 ) 푙95− 푙 푃(푙) = 푃푚푎푥 (1 + ℯ 50 )

where P (l) is the proportion mature at TL, l and Pmax is the maximum proportion of mature individuals. The lengths of which 50 and 95% of the population are mature (l50 and l95) were estimated using a generalised linear model (GLM) with a binomial error structure and a logit- link function using the ‘psyphy’ package version 0.1-9 (Knoblauch, 2014) and the ‘FSA’ package version 0.0.11 (Ogle, 2016) in the R program environment version 3.2.2 (R Core

Team, 2015). Age-at-maturity was determined by substituting length with age. A50 and A95 were the ages at which 50 and 95% of the population reached maturity.

4.3 Results

4.3.1 Age determination

In total 186 individuals were collected: 131 females and 55 males. Males ranged in size from

31–53 cm (TL) and females from 31–66 cm (TL). The majority of sharks were aged between

0 and 1 years (i.e. birthmark was present but not fully formed 1st band pair) (Fig 4-2). Final partial ages ranged from 0.2 to 4.6 years. The oldest female was 64 cm (TL) and aged at 4.6 years. The oldest male was 51 cm (TL) and aged at 3.6 years.

The measures of variability around the determination of ages were high compared to other elasmobranch ageing studies (Campana, 2001, Natanson et al., 2007, Gutteridge et al., 2013).

The Average Percent Error (APE), Chang’s CV and PA ± 1 year were 29.1%, 41.1% and

62.4%, respectively. Higher variability will be experienced when ageing short lived species as small differences in band pair counts can produce inflated error estimates in comparison to longer lived species (Simpfendorfer, 1993). Bowker’s test for symmetry (df = 8, x2 = 16.4, P

= 0.037) indicated some systematic bias between readers. The age bias plot (Fig 4-3) showed that this bias was associated with reader 1 estimating younger counts of band pairs at 3 and 4 years relative to reader 2. The use of consensus counts to produce final ages overcame this ageing bias.

Figure 4-2: Frequency histogram of samples for each age class.

Figure 4-3: Age bias plot showing agreement between two independent readers. The PA ± 1 year was 62.4%, APE was 29.1 and Chang’s coefficient of variation (CV) was 41.1%.

4.3.2 Growth model fitting

Without data from small new born animals three-parameter models were unsuitable as the projected length-at-birth values were too high and biologically unreasonable for R. taylori

(37– 38 cm) (Table 4-3).The three-parameter models with the four added size at birth individuals had similar AIC weights for combined and individual sexes (Table 4-4). All three candidate models had similar weights in the three-parameter models. Neither of the three- parameter approaches accurately represented the early growth of R. taylori, over-estimating

66 the size at birth. Amongst the two-parameter models the VBGM performed best as neither logistic and Gompertz models had ∆ values < 2, although there was some weak support for the Gompertz model for males (w = 0.24) (Table 4-5). The two-parameter models projected much higher growth completion rates (k, g(log), g(gom)) than three-parameter models however, the fixed length-at-birth value were more realistic. Thus, it is likely that none of the fitting approaches produced accurate estimates of all three parameters. However, the two-parameter

VBGM is recommended to describe the growth of R. taylori in the GoP (Fig 4-4), with a growth estimate (k) of 1.27 for both sexes combined (Table 4-5). A likelihood ratio test showed significant difference (df = 3, x2 = 23.3, P = 3.5) in the VBGM fit between males and females which demonstrated that sexes should be modelled separately. The error estimates for the male VBGM parameters were much higher than for females, indicating much greater level of uncertainty, probably because of the smaller sample size.

Table 4-3: Summary of results from the multi model approach incorporating Akaike’s information Criterion (AIC) using three-parameter versions of models.

Sex Model n AICC ∆ W L0(±SE) L∞(±SE) k(±SE) g(log)(±SE) g(gom)(±SE) RSE (%)

Combined VB3 186 1129.06 0.53 0.29 37.89±1.27 74.34±12.98 0.25±0.14 4.96

Logistic 186 1128.53 0 0.38 38.17±1.11 66.92±6.0 0.50±0.14 4.96

Gompertz 186 1128.78 0.25 0.33 38.03±1.18 69.65±8.21 0.38±0.14 4.96

Male VB3 55 306.3 0.17 0.32 38.48±1.50 58.89±15.72 0.31±0.37 3.72

Logistic 55 306.13 0 0.35 38.51±0.76 55.71±8.90 0.51±0.20 3.71

Gompertz 55 306.22 0.09 0.33 38.50±1.44 57.00±11.41 0.41±0.37 3.72

Female VB3 131 801.08 0.29 0.31 38.03±1.90 71.08±10.55 0.31±0.17 5.04 Logistic 131 800.8 0 0.36 38.53±1.35 66.30±5.79 0.55±0.15 5.04

Gompertz 131 800.93 0.13 0.33 38.30±1.74 68.17±7.46 0.43±0.17 5.04

n is the sample size, AICC is the small-sample bias adjusted from the Akaike’s Information Criteria, ∆ is the difference in AICC values between models, w (%) are the AICC weights, L0 and L∞ are the length-at-birth and asymptotic length in cm respectively, k is the growth completion rate

-1 in (year ) for the VB3, g(log) and g(gom) are the growth parameters for Logistic and Gompertz functions respectively, SE is the standard error of each growth parameter and RSE is the residual standard error for the models.

Table 4-4: Summary of results from the multi model approach incorporating Akaike’s information Criterion (AIC) using three-parameter versions of models with four hypothetical aged zero individuals.

Sex Model n AICC ∆ W L0(±SE) L∞(±SE) k(±SE) g(log)(±SE) g(gom)(±SE) RSE (%)

Combined VB3 190 1166.85 0 0.45 35.12±1.32 63.88±4.03 0.48±0.14 5.15

Logistic 190 1168.21 1.96 0.23 35.98±1.14 61.75±2.87 0.73±0.15 5.16

Gompertz 190 1167.59 0.73 0.32 35.59±1.22 62.65±3.33 0.60±0.14 5.16

Male VB3 57 330.66 0 0.39 34.55±1.87 50.42±2.57 1.01±0.43 4.19

Logistic 57 331.28 0.62 0.28 35.28±0.92 50.41±2.47 1.17±0.25 4.21

Gompertz 57 331.01 0.35 0.33 34.96±1.76 50.44±2.53 1.08±0.44 4.2

Female VB3 133 819.85 0 0.44 34.91±1.96 63.77±3.92 0.53±0.17 5.17 Logistic 133 821.06 1.21 0.24 36.22±1.38 62.27±3.04 0.77±0.15 5.20

Gompertz 133 820.51 0.66 0.32 35.64±1.8 62.92±3.41 0.65±0.18 5.19

69 n is the sample size, AICC is the small-sample bias adjusted from the Akaike’s Information Criteria, ∆ is the difference in AICC values between models, w (%) are the AICC weights, L0 and L∞ are the length-at-birth and asymptotic length in cm respectively, k is the growth completion rate

Sex Model n AICC ∆ W L∞(±SE) k(±SE) g(log)(±SE) g(gom)(±SE) RSE (%)

Combined VB2 186 1193.71 0 0.99 55.95±0.95 1.27±0.11 5.54

Logistic 186 1213.08 19.38 0 54.41±0.75 2.12±0.14 5.83

Gompertz 186 1203.61 9.9 0.01 55.07±0.82 1.67±0.13 5.68

Male VB2 55 336.13 0 0.64 46.11±0.9 3.69±0.68 4.44 Logistic 55 339.47 3.34 0.12 45.08±0.77 6.73±1.23 4.57

Gompertz 55 338.1 1.97 0.24 45.52±0.82 5.04±0.92 4.52

Female VB2 131 830.37 0 0.96 57.78±1.12 1.17±0.12 5.40

Logistic 131 842.88 12.52 0.00 56.08±0.84 1.98±0.15 5.66

Gompertz 131 836.6 6.23 0.04 56.8±0.94 1.55±0.13 5.53

n is the sample size, AICC is the small-sample bias adjusted from the Akaike’s Information Criteria, ∆ is the difference in AICC values between

-1 models, w (%) ar